chore: release v2.61.0

move strict to types instead of test/middleware code

.devcontainer/devcontainer.json

@@ -0,0 +1,14 @@
+{
+	"name": "langfuse dev",
+    "image": "mcr.microsoft.com/devcontainers/javascript-node:20-bookworm",
+    "features": {
+        "ghcr.io/devcontainers/features/docker-in-docker:2": {
+        "version": "latest",
+        "moby": true
+        }
+    },
+    "forwardPorts": [3000],
+
+    "onCreateCommand": "npm install -g pnpm@8.15.5",
+    "postCreateCommand": "cp .env.dev.example .env && pnpm run dx-f"
+}

.env.dev.example

@@ -34,10 +34,6 @@ S3_SECRET_ACCESS_KEY=
 S3_BUCKET_NAME=
 S3_REGION=
 
-# Prompt playground
-OPENAI_API_KEY=""
-ANTHROPIC_API_KEY=""
-
 # Set during docker build of application
 # Used to disable environment verification at build time
 # DOCKER_BUILD=1

.env.local.example

@@ -13,11 +13,6 @@ NEXTAUTH_URL="http://localhost:3000"
 LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES="false"
 SALT="salt"
 
-
-# Prompt playground
-OPENAI_API_KEY=""
-ANTHROPIC_API_KEY=""
-
 # Redis
 REDIS_HOST="127.0.0.1"
 REDIS_PORT=6379

.env.prod.example

@@ -9,6 +9,8 @@
 DATABASE_URL="postgresql://postgres:postgres@db:5432/postgres"
 # DIRECT_URL="postgresql://postgres:postgres@db:5432/postgres"
 # SHADOW_DATABASE_URL=
+# optional, set to true to disable automated database migrations on Docker start
+# LANGFUSE_AUTO_POSTGRES_MIGRATION_DISABLED= 
 
 # Next Auth
 # NEXTAUTH_URL does not need to be set when deploying on Vercel
@@ -23,6 +25,7 @@ SALT="salt"
 # Use CSP headers to enforce HTTPS, optional
 LANGFUSE_CSP_ENFORCE_HTTPS="true"
 
+
 # Docker only, optional
 # PORT=3000
 # HOSTNAME=localhost
@@ -34,6 +37,9 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # Enable experimental features, optional
 # LANGFUSE_ENABLE_EXPERIMENTAL_FEATURES=true
 
+# Disable automated database migrations on Docker start, optional
+# LANFUSE_WEB_MIGRATION_DISABLED=true
+
 # Auth, optional configuration
 # AUTH_DOMAINS_WITH_SSO_ENFORCEMENT=domain1.com,domain2.com
 # AUTH_DISABLE_USERNAME_PASSWORD=true
@@ -43,6 +49,7 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # AUTH_GOOGLE_CLIENT_ID=
 # AUTH_GOOGLE_CLIENT_SECRET=
 # AUTH_GOOGLE_ALLOW_ACCOUNT_LINKING=false
+# AUTH_GOOGLE_ALLOWED_DOMAINS=langfuse.com,google.com # optional allowlist of workspace domains that can sign in via Google
 # AUTH_GITHUB_CLIENT_ID=
 # AUTH_GITHUB_CLIENT_SECRET=
 # AUTH_GITHUB_ALLOW_ACCOUNT_LINKING=false
@@ -62,7 +69,12 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # AUTH_COGNITO_CLIENT_SECRET=
 # AUTH_COGNITO_ISSUER=
 # AUTH_COGNITO_ALLOW_ACCOUNT_LINKING=false
-
+# AUTH_CUSTOM_CLIENT_ID=
+# AUTH_CUSTOM_CLIENT_SECRET=
+# AUTH_CUSTOM_ISSUER=
+# AUTH_CUSTOM_NAME=
+# AUTH_CUSTOM_SCOPE="openid email profile" # optional
+# AUTH_CUSTOM_ALLOW_ACCOUNT_LINKING=false
 
 # Transactional email, optional
 # Defines the email address to use as the from address.
@@ -76,6 +88,7 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # S3_SECRET_ACCESS_KEY=
 # S3_BUCKET_NAME=
 # S3_REGION=
+# BATCH_EXPORT_DOWNLOAD_LINK_EXPIRATION_HOURS=
 
 # Exports are streamed to S3 in pages to avoid memory issues
 # The page size can be adjusted if needed to optimize performance
@@ -118,18 +131,14 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # NEXT_SENTRY_ORG=
 # NEXT_SENTRY_PROJECT=
 # SENTRY_AUTH_TOKEN=
+# Used to determine the Sentry sample rate
+# LANGFUSE_TRACING_SAMPLE_RATE=
+
 
 # Cloudflare Turnstile
 # NEXT_PUBLIC_TURNSTILE_SITE_KEY=
 # TURNSTILE_SECRET_KEY=
 
-# Prompt playground
-# OPENAI_API_KEY=""
-# ANTHROPIC_API_KEY=""
-
-# Betterstack
-# LANGFUSE_TEAM_BETTERSTACK_TOKEN=
-
 # Demo project that users can use to try the platform
 # NEXT_PUBLIC_DEMO_PROJECT_ID=
 
@@ -142,4 +151,5 @@ LANGFUSE_CSP_ENFORCE_HTTPS="true"
 # Admin API
 # ADMIN_API_KEY=
 
+
 ### END Langfuse Cloud Config

.github/workflows/pipeline.yml

@@ -40,25 +40,35 @@ jobs:
       - name: Checkout
         uses: actions/checkout@v3
 
-      - name: Build Web Docker image
-        uses: docker/build-push-action@v4
-        with:
-          context: .
-          file: ./web/Dockerfile
-          push: false
+      - name: Build and run both images from compose
+        run: |
+          docker compose -f docker-compose.build.yml up -d
+          sleep 5 # Wait for PostgreSQL to accept connections
 
-      - name: Build Worker Docker image
-        uses: docker/build-push-action@v4
-        with:
-          context: .
-          file: ./worker/Dockerfile
-          push: false
+      - name: Check server health
+        run: |
+          timeout 10 bash -c 'until curl -f http://localhost:3000/api/public/health; do sleep 2; done'
+
+      - name: Check worker health
+        run: |
+          timeout 10 bash -c 'until curl -f http://localhost:3030/api/health; do sleep 2; done'
+
+      - name: Ensure no unhealthy status
+        run: |
+          if docker-compose ps | grep "(unhealthy)"; then
+            echo "One or more services are unhealthy"
+            exit 1
+          else
+            echo "All services are healthy"
+          fi
 
   tests-web:
     runs-on: ubuntu-latest
+    name: tests-web (node${{ matrix.node-version }}, pg${{ matrix.postgres-version }})
     strategy:
       matrix:
         node-version: [20]
+        postgres-version: [12, 15]
     steps:
       - name: Set Swap Space
         uses: pierotofy/set-swap-space@master
@@ -89,6 +99,9 @@ jobs:
         run: |
           docker compose -f docker-compose.dev.yml up -d
           sleep 5 # Wait for PostgreSQL to accept connections
+          docker compose ps
+        env:
+          POSTGRES_VERSION: ${{ matrix.postgres-version }}
 
       - name: Seed DB
         run: |
@@ -106,9 +119,11 @@ jobs:
 
   tests-worker:
     runs-on: ubuntu-latest
+    name: tests-worker (node${{ matrix.node-version }}, pg${{ matrix.postgres-version }})
     strategy:
       matrix:
         node-version: [20]
+        postgres-version: [12, 15]
     steps:
       - name: Set Swap Space
         uses: pierotofy/set-swap-space@master
@@ -139,6 +154,7 @@ jobs:
         run: |
           docker compose -f docker-compose.dev.yml up -d
           sleep 5 # Wait for PostgreSQL to accept connections
+          docker compose ps
 
       - name: Seed DB
         run: |
@@ -176,6 +192,7 @@ jobs:
       - name: Run + migrate
         run: |
           docker compose -f docker-compose.dev.yml up -d
+          docker compose ps
           sleep 5 # Wait for PostgreSQL to accept connections
 
       - name: Seed DB
@@ -242,6 +259,14 @@ jobs:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
 
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+        with:
+          driver-opts: network=host
+
       - name: Extract metadata (tags, labels) for Docker
         id: meta-web
         uses: docker/metadata-action@v4
@@ -265,6 +290,9 @@ jobs:
           push: true
           tags: ${{ steps.meta-web.outputs.tags }}
           labels: ${{ steps.meta-web.outputs.labels }}
+          platforms: |
+            linux/amd64
+            ${{ startsWith(github.ref, 'refs/tags/') && 'linux/arm64' || '' }}
 
       - name: Extract metadata (tags, labels) for Docker
         id: meta-worker
@@ -289,3 +317,6 @@ jobs:
           push: true
           tags: ${{ steps.meta-worker.outputs.tags }}
           labels: ${{ steps.meta-worker.outputs.labels }}
+          platforms: |
+            linux/amd64
+            ${{ startsWith(github.ref, 'refs/tags/') && 'linux/arm64' || '' }}

.github/workflows/porter_stack_web_eu_temp.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+    - production
+name: Deploy to web
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v3
+    - name: Set Github tag
+      id: vars
+      run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+    - name: Setup porter
+      uses: porter-dev/setup-porter@v0.1.0
+    - name: Deploy stack
+      timeout-minutes: 30
+      run: exec porter apply
+      env:
+        PORTER_CLUSTER: "4060"
+        PORTER_DEPLOYMENT_TARGET_ID: 3be075c0-b121-4123-aec1-5dc571626b0c
+        PORTER_HOST: https://dashboard.porter.run
+        PORTER_PR_NUMBER: ${{ github.event.number }}
+        PORTER_PROJECT: "12565"
+        PORTER_REPO_NAME: ${{ github.event.repository.name }}
+        PORTER_STACK_NAME: web
+        PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+        PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4060 }}

.github/workflows/porter_stack_web_staging.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+      - main
+name: Deploy to web (Staging)
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4037"
+          PORTER_DEPLOYMENT_TARGET_ID: c22d8fa2-f97c-4e1f-9cae-83cc6bfbad35
+          PORTER_HOST: https://dashboard.getporter.dev
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: web
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4037 }}

.github/workflows/porter_stack_web_us_temp.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+      - production
+name: Deploy to web (US temp)
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4054"
+          PORTER_DEPLOYMENT_TARGET_ID: cddc0f2a-bb19-401a-9819-29f59f76b101
+          PORTER_HOST: https://dashboard.porter.run
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: web
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4054 }}

.github/workflows/porter_stack_worker_eu_temp.yml

@@ -1,8 +1,8 @@
 "on":
   push:
     branches:
-    - main
-name: Deploy to worker
+    - production
+name: Deploy to worker (EU temp)
 jobs:
   porter-deploy:
     runs-on: ubuntu-latest
@@ -18,12 +18,12 @@ jobs:
       timeout-minutes: 30
       run: exec porter apply
       env:
-        PORTER_CLUSTER: "3959"
-        PORTER_DEPLOYMENT_TARGET_ID: d2bb23e4-1d77-48f2-a113-383d076959b3
-        PORTER_HOST: https://dashboard.getporter.dev
+        PORTER_CLUSTER: "4060"
+        PORTER_DEPLOYMENT_TARGET_ID: 3be075c0-b121-4123-aec1-5dc571626b0c
+        PORTER_HOST: https://dashboard.porter.run
         PORTER_PR_NUMBER: ${{ github.event.number }}
         PORTER_PROJECT: "12565"
         PORTER_REPO_NAME: ${{ github.event.repository.name }}
         PORTER_STACK_NAME: worker
         PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
-        PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_3959 }}
+        PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4060 }}

.github/workflows/porter_stack_worker_staging.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+      - main
+name: Deploy to worker (staging)
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4037"
+          PORTER_DEPLOYMENT_TARGET_ID: c22d8fa2-f97c-4e1f-9cae-83cc6bfbad35
+          PORTER_HOST: https://dashboard.getporter.dev
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: worker
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4037 }}

.github/workflows/porter_stack_worker_us.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+      - production
+name: Deploy to worker (US)
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4038"
+          PORTER_DEPLOYMENT_TARGET_ID: 260cd04e-5dc8-4b22-a291-ad4453f981f7
+          PORTER_HOST: https://dashboard.getporter.dev
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: worker
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4038 }}

.github/workflows/porter_stack_worker_us_temp.yml

@@ -0,0 +1,29 @@
+"on":
+  push:
+    branches:
+      - production
+name: Deploy to worker (US temp)
+jobs:
+  porter-deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Set Github tag
+        id: vars
+        run: echo "sha_short=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+      - name: Setup porter
+        uses: porter-dev/setup-porter@v0.1.0
+      - name: Deploy stack
+        timeout-minutes: 30
+        run: exec porter apply
+        env:
+          PORTER_CLUSTER: "4054"
+          PORTER_DEPLOYMENT_TARGET_ID: cddc0f2a-bb19-401a-9819-29f59f76b101
+          PORTER_HOST: https://dashboard.porter.run
+          PORTER_PR_NUMBER: ${{ github.event.number }}
+          PORTER_PROJECT: "12565"
+          PORTER_REPO_NAME: ${{ github.event.repository.name }}
+          PORTER_STACK_NAME: worker
+          PORTER_TAG: ${{ steps.vars.outputs.sha_short }}
+          PORTER_TOKEN: ${{ secrets.PORTER_STACK_12565_4054 }}

.github/workflows/snyk.yml

@@ -0,0 +1,39 @@
+name: Snyk Container
+on: push
+jobs:
+  snyk:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - name: Build a Docker image
+        run: docker compose -f docker-compose.build.yml up -d
+
+      - name: Run Snyk to check Docker image for vulnerabilities (langfuse-server)
+        continue-on-error: true
+        uses: snyk/actions/docker@master
+        env:
+          SNYK_TOKEN: ${{ secrets.SNYK_TOKEN }}
+        with:
+          image: langfuse-server
+          args: --file=web/Dockerfile
+
+      - name: Upload result to GitHub Code Scanning
+        uses: github/codeql-action/upload-sarif@v2
+        with:
+          sarif_file: snyk.sarif
+          category: web
+
+      - name: Run Snyk to check Docker image for vulnerabilities (langfuse-worker)
+        continue-on-error: true
+        uses: snyk/actions/docker@master
+        env:
+          SNYK_TOKEN: ${{ secrets.SNYK_TOKEN }}
+        with:
+          image: langfuse-worker
+          args: --file=worker/Dockerfile
+
+      - name: Upload result to GitHub Code Scanning
+        uses: github/codeql-action/upload-sarif@v2
+        with:
+          sarif_file: snyk.sarif
+          category: worker

.husky/pre-commit

@@ -0,0 +1,22 @@
+#!/bin/sh
+
+# Get the current branch
+current_branch=$(git rev-parse --abbrev-ref HEAD)
+
+# Define the protected branch
+protected_branch="main"
+
+# Check if the current branch is the protected branch
+if [ "$current_branch" = "$protected_branch" ]; then
+    echo "🚨 You are about to commit to the $protected_branch branch. Are you sure? (y/n)"
+    read -r answer < /dev/tty
+    if [ "$answer" != "${answer#[Yy]}" ]; then
+        exit 0 # Commit will proceed
+    else
+        echo "Commit to $protected_branch branch has been canceled."
+        exit 1 # Commit will be blocked
+    fi
+fi
+
+# If not the protected branch, proceed with the commit
+exit 0

.husky/pre-push

@@ -0,0 +1,22 @@
+#!/bin/sh
+
+# Get the current branch
+current_branch=$(git rev-parse --abbrev-ref HEAD)
+
+# Define the protected branch
+protected_branch="main"
+
+# Check if the current branch is the protected branch
+if [ "$current_branch" = "$protected_branch" ]; then
+    echo "🚨 You are about to push to the $protected_branch branch. Are you sure? (y/n)"
+    read -r answer < /dev/tty
+    if [ "$answer" != "${answer#[Yy]}" ]; then
+        exit 0 # Push will proceed
+    else
+        echo "Push to $protected_branch branch has been canceled."
+        exit 1 # Push will be blocked
+    fi
+fi
+
+# If not the protected branch, proceed with the push
+exit 0

CONTRIBUTING.md

@@ -55,7 +55,7 @@ A good first step is to search for open [issues](https://github.com/langfuse/lan
 
 > [!NOTE]
 > Infrastructure will change in Langfuse version 3.0. More in the [GitHub Discussions](https://github.com/orgs/langfuse/discussions/1902).
-> `langfuse/langfuse/worker` is under active development and not recommended for production use in Langfuse 2.x
+> `langfuse/langfuse/worker` is under active development and not recommended for production use in Langfuse 2.x.
 
 ```mermaid
 flowchart TB
@@ -151,6 +151,8 @@ Requirements
 - Node.js 20 as specified in the [.nvmrc](.nvmrc)
 - Docker to run the database locally
 
+**Note:** You can also simply run Langfuse in a **GitHub Codespace** via the provided devcontainer. To do this, click on the green "Code" button in the top right corner of the repository and select "Open with Codespaces".
+
 **Steps**
 
 1. Fork the repository and clone it locally
@@ -191,6 +193,16 @@ Requirements
     pnpm run dev
    ```
 
+7. Open the web app in the browser:
+
+   http://localhost:3000
+
+8. Log in as a test user (after you ran `db:seed` command):
+
+   Username: demo@langfuse.com
+
+   Password: password
+
 ## Monorepo quickstart
 
 - Available packages and their dependencies
@@ -358,7 +370,9 @@ The background color of the following component will be `hsl(var(--primary))` an
 
 By following these guidelines, you can ensure that any contributions to our theme are consistent, maintainable, and aligned with our design system.
 
-## Using secrets stored in 1Password
+## Maintainers
+
+### Using secrets stored in 1Password
 
 When applying changes to non-local environments, you may need to use secrets stored in 1Password. We use the 1Password CLI for this purpose.
 

README.md

@@ -24,6 +24,8 @@
    </div>
    <span>Langfuse uses <a href="https://github.com/orgs/langfuse/discussions"><strong>Github Discussions</strong></a>  for Support and Feature Requests.</span>
    <br/>
+   <span>We're hiring. <a href="https://langfuse.com/careers"><strong>Join us</strong></a> in Backend Engineering, Product Engineering, and Developer Relations.</span>
+   <br/>
    <br/>
    <div>
       <a href="https://github.com/langfuse/langfuse/blob/main/LICENSE"><img src="https://img.shields.io/badge/License-MIT-red.svg?style=flat-square" alt="MIT License"></a>
@@ -157,7 +159,7 @@ This repository is MIT licensed, except for the `ee` folders. See [LICENSE](LICE
 
 ### GET API to export your data
 
-[**GET routes**](https://langfuse.com/docs/integrations/api) to use data in downstream applications (e.g. embedded analytics).
+[**GET routes**](https://langfuse.com/docs/integrations/api) to use data in downstream applications (e.g. embedded analytics). You can also access them conveniently via the SDKs ([docs](https://langfuse.com/docs/query-traces)).
 
 ### Security & Privacy
 

docker-compose.build.yml

@@ -5,6 +5,8 @@ services:
     build:
       dockerfile: ./web/Dockerfile
       context: .
+      args:
+        - NEXT_PUBLIC_LANGFUSE_CLOUD_REGION=${NEXT_PUBLIC_LANGFUSE_CLOUD_REGION}
     depends_on:
       - db
       - redis
@@ -20,11 +22,18 @@ services:
       - LANGFUSE_WORKER_HOST=${LANGFUSE_WORKER_HOST:-worker}
       - LANGFUSE_WORKER_PASSWORD=${LANGFUSE_WORKER_PASSWORD:-mybasicauthsecret}
     restart: always
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:3000/api/public/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
 
   worker:
     build:
       dockerfile: ./worker/Dockerfile
       context: .
+      args:
+        - NEXT_PUBLIC_LANGFUSE_CLOUD_REGION=${NEXT_PUBLIC_LANGFUSE_CLOUD_REGION}
     depends_on:
       - db
       - redis
@@ -41,6 +50,11 @@ services:
       - REDIS_AUTH=${REDIS_AUTH:-myredissecret}
       - LANGFUSE_WORKER_PASSWORD=${LANGFUSE_WORKER_PASSWORD:-mybasicauthsecret}
     restart: always
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:3030/api/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
 
   redis:
     image: redis:7.2.4

docker-compose.dev.yml

@@ -1,5 +1,3 @@
-version: "3.5"
-
 services:
   redis:
     image: redis:7.2.4
@@ -10,7 +8,7 @@ services:
       - 6379:6379
 
   db:
-    image: postgres
+    image: postgres:${POSTGRES_VERSION:-latest}
     restart: always
     command: ["postgres", "-c", "log_statement=all"]
     environment:

docker-compose.yml

@@ -1,8 +1,6 @@
-version: "3.5"
-
 services:
   langfuse-server:
-    image: ghcr.io/langfuse/langfuse:latest
+    image: langfuse/langfuse:2
     depends_on:
       db:
         condition: service_healthy

ee/package.json

@@ -27,13 +27,13 @@
     "@langfuse/shared": "workspace:*",
     "next": "^14.2.3",
     "next-auth": "^4.24.7",
-    "zod": "^3.22.4"
+    "zod": "^3.23.8"
   },
   "devDependencies": {
     "@repo/eslint-config": "*",
     "@repo/typescript-config": "*",
     "@types/node": "^20.11.29",
-    "@typescript-eslint/parser": "^7.7.0",
+    "@typescript-eslint/parser": "^7.12.0",
     "eslint": "^8.57.0",
     "eslint-config-prettier": "^9.1.0",
     "eslint-config-standard": "^17.1.0",

ee/src/sso/index.ts

@@ -11,6 +11,7 @@ import { encrypt, decrypt } from "@langfuse/shared/encryption";
 import { SsoProviderSchema } from "./types";
 import { type NextApiRequest, type NextApiResponse } from "next";
 import { env } from "../env";
+import { CustomSSOProvider } from "@langfuse/shared/src/server";
 
 // Local cache for SSO configurations
 let cachedSsoConfigs: {
@@ -173,6 +174,15 @@ const dbToNextAuthProvider = (provider: SsoProviderSchema): Provider | null => {
       ...provider.authConfig,
       clientSecret: decrypt(provider.authConfig.clientSecret),
     });
+  else if (provider.authProvider === "custom")
+    return CustomSSOProvider({
+      id: getAuthProviderIdForSsoConfig(provider), // use the domain as the provider id as we use domain-specific credentials
+      ...provider.authConfig,
+      clientSecret: decrypt(provider.authConfig.clientSecret),
+      authorization: {
+        params: { scope: provider.authConfig.scope ?? "openid email profile" },
+      },
+    });
   else {
     // Type check to ensure we handle all providers
     // eslint-disable-next-line no-unused-vars

ee/src/sso/types.ts

@@ -76,12 +76,27 @@ export const CognitoProviderSchema = base.extend({
     .nullish(),
 });
 
+export const CustomProviderSchema = base.extend({
+  authProvider: z.literal("custom"),
+  authConfig: z
+    .object({
+      name: z.string(),
+      clientId: z.string(),
+      clientSecret: z.string(),
+      issuer: z.string(),
+      scope: z.string().nullish(),
+      allowDangerousEmailAccountLinking: z.boolean().optional().default(false),
+    })
+    .nullish(),
+});
+
 export type GoogleProviderSchema = z.infer<typeof GoogleProviderSchema>;
 export type GithubProviderSchema = z.infer<typeof GithubProviderSchema>;
 export type Auth0ProviderSchema = z.infer<typeof Auth0ProviderSchema>;
 export type OktaProviderSchema = z.infer<typeof OktaProviderSchema>;
 export type AzureAdProviderSchema = z.infer<typeof AzureAdProviderSchema>;
 export type CognitoProviderSchema = z.infer<typeof CognitoProviderSchema>;
+export type CustomProviderSchema = z.infer<typeof CustomProviderSchema>;
 
 export const SsoProviderSchema = z.discriminatedUnion("authProvider", [
   GoogleProviderSchema,
@@ -90,6 +105,7 @@ export const SsoProviderSchema = z.discriminatedUnion("authProvider", [
   OktaProviderSchema,
   AzureAdProviderSchema,
   CognitoProviderSchema,
+  CustomProviderSchema,
 ]);
 
 export type SsoProviderSchema = z.infer<typeof SsoProviderSchema>;

fern/apis/client/definition/score.yml

@@ -9,7 +9,6 @@ service:
       method: POST
       path: /scores
       request: CreateScoreRequest
-      response: Score
 
 types:
   CreateScoreRequest:
@@ -17,15 +16,119 @@ types:
       id: optional<string>
       traceId: string
       name: string
-      value: double
+      value:
+        type: CreateScoreValue
+        docs: The value of the score. Must be passed as string for categorical scores, and numeric for boolean and numeric scores. Boolean score values must equal either 1 or 0 (true or false)
       observationId: optional<string>
       comment: optional<string>
-  Score:
+      dataType:
+        type: optional<ScoreDataType>
+        docs: When set, must match the score value's type. If not set, will be inferred from the score value or config
+      configId:
+        type: optional<string>
+        docs: Reference a score config on a score. When set, the score name must equal the config name and scores must comply with the config's range and data type. For categorical scores, the value must map to a config category. Numeric scores might be constrained by the score config's max and min values
+    examples:
+      - value:
+          name: "novelty"
+          value: 0.9
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "consistency"
+          value: 1.2
+          dataType: "NUMERIC"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "accuracy"
+          value: 0.9
+          dataType: "NUMERIC"
+          configId: "9203-4567-89ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "toxicity"
+          value: "not toxic"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "correctness"
+          value: "partially correct"
+          dataType: "CATEGORICAL"
+          configId: "1234-5678-90ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "hallucination"
+          value: 0
+          dataType: "BOOLEAN"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "helpfulness"
+          value: 1
+          dataType: "BOOLEAN"
+          configId: "1234-5678-90ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+  BaseScore:
     properties:
       id: string
       traceId: string
       name: string
-      value: double
+      source: ScoreSource
       observationId: optional<string>
       timestamp: datetime
+      createdAt: datetime
+      updatedAt: datetime
+      authorUserId: optional<string>
       comment: optional<string>
+      configId:
+        type: optional<string>
+        docs: Reference a score config on a score. When set, config and score name must be equal and value must comply to optionally defined numerical range
+  NumericScore:
+    extends: BaseScore
+    properties:
+      value:
+        type: double
+        docs: The numeric value of the score
+  BooleanScore:
+    extends: BaseScore
+    properties:
+      value:
+        type: double
+        docs: The numeric value of the score. Equals 1 for "True" and 0 for "False"
+      stringValue:
+        type: string
+        docs: The string representation of the score value. Is inferred from the numeric value and equals "True" or "False"
+  CategoricalScore:
+    extends: BaseScore
+    properties:
+      value:
+        type: optional<double>
+        docs: Only defined if a config is linked. Represents the numeric category mapping of the stringValue
+      stringValue:
+        type: string
+        docs: The string representation of the score value. If no config is linked, can be any string. Otherwise, must map to a config category
+  Score:
+    discriminant: "dataType"
+    union:
+      NUMERIC:
+        type: NumericScore
+        docs: "Score with NUMERIC data type"
+      CATEGORICAL:
+        type: CategoricalScore
+        docs: "Score with CATEGORICAL data type"
+      BOOLEAN:
+        type: BooleanScore
+        docs: "Score with BOOLEAN data type"
+  ScoreSource:
+    enum:
+      - ANNOTATION
+      - API
+      - EVAL
+  ScoreDataType:
+    enum:
+      - NUMERIC
+      - CATEGORICAL
+      - BOOLEAN
+
+  CreateScoreValue:
+    discriminated: false
+    union:
+      - double
+      - string
+    docs: The value of the score. Must be passed as string for categorical scores, and numeric for boolean and numeric scores

fern/apis/server/definition/commons.yml

@@ -42,11 +42,18 @@ types:
       htmlPath:
         type: string
         docs: Path of trace in Langfuse UI
+      latency:
+        type: double
+        docs: Latency of trace in seconds
       totalCost:
         type: double
         docs: Cost of trace in USD
-      observations: list<ObservationsView>
-      scores: list<Score>
+      observations:
+        type: list<ObservationsView>
+        docs: List of observations
+      scores:
+        type: list<Score>
+        docs: List of scores
   Session:
     properties:
       id: string
@@ -88,6 +95,7 @@ types:
       calculatedOutputCost: optional<double>
       calculatedTotalCost: optional<double>
       latency: optional<double>
+      timeToFirstToken: optional<double>
 
   Usage:
     docs: Standard interface for usage and cost
@@ -111,18 +119,91 @@ types:
       totalCost:
         docs: USD total cost, defaults to input+output
         type: optional<double>
-
-  Score:
+  ScoreConfig:
+    docs: Configuration for a score
+    properties:
+      id: string
+      name: string
+      createdAt: datetime
+      updatedAt: datetime
+      projectId: string
+      dataType: ScoreDataType
+      isArchived:
+        type: boolean
+        docs: Whether the score config is archived. Defaults to false
+      minValue:
+        type: optional<double>
+        docs: Sets minimum value for numerical scores. If not set, the minimum value defaults to -∞
+      maxValue:
+        type: optional<double>
+        docs: Sets maximum value for numerical scores. If not set, the maximum value defaults to +∞
+      categories:
+        type: optional<list<ConfigCategory>>
+        docs: Configures custom categories for categorical scores
+      description: optional<string>
+  ConfigCategory:
+    properties:
+      value: double
+      label: string
+  BaseScore:
     properties:
       id: string
       traceId: string
       name: string
-      value: double
       source: ScoreSource
       observationId: optional<string>
       timestamp: datetime
+      createdAt: datetime
+      updatedAt: datetime
+      authorUserId: optional<string>
       comment: optional<string>
-  DatasetCore:
+      configId:
+        type: optional<string>
+        docs: Reference a score config on a score. When set, config and score name must be equal and value must comply to optionally defined numerical range
+  NumericScore:
+    extends: BaseScore
+    properties:
+      value:
+        type: double
+        docs: The numeric value of the score
+  BooleanScore:
+    extends: BaseScore
+    properties:
+      value:
+        type: double
+        docs: The numeric value of the score. Equals 1 for "True" and 0 for "False"
+      stringValue:
+        type: string
+        docs: The string representation of the score value. Is inferred from the numeric value and equals "True" or "False"
+  CategoricalScore:
+    extends: BaseScore
+    properties:
+      value:
+        type: optional<double>
+        docs: Only defined if a config is linked. Represents the numeric category mapping of the stringValue
+      stringValue:
+        type: string
+        docs: The string representation of the score value. If no config is linked, can be any string. Otherwise, must map to a config category
+  Score:
+    discriminant: "dataType"
+    union:
+      NUMERIC:
+        type: NumericScore
+        docs: "Score with NUMERIC data type"
+      CATEGORICAL:
+        type: CategoricalScore
+        docs: "Score with CATEGORICAL data type"
+      BOOLEAN:
+        type: BooleanScore
+        docs: "Score with BOOLEAN data type"
+
+  CreateScoreValue:
+    discriminated: false
+    union:
+      - double
+      - string
+    docs: The value of the score. Must be passed as string for categorical scores, and numeric for boolean and numeric scores
+  Dataset:
     properties:
       id: string
       name: string
@@ -131,21 +212,6 @@ types:
       projectId: string
       createdAt: datetime
       updatedAt: datetime
-  DatasetWithReferences: # GET datasets
-    extends: DatasetCore
-    properties:
-      items:
-        type: list<string>
-        docs: list of dataset item ids
-      runs:
-        type: list<string>
-        docs: list of dataset run names
-  Dataset: # GET datasets/[name]
-    docs: Dataset including all items
-    extends: DatasetCore
-    properties:
-      items: list<DatasetItem>
-      runs: list<string>
   DatasetItem:
     properties:
       id: string
@@ -179,7 +245,43 @@ types:
       datasetName: string
       createdAt: datetime
       updatedAt: datetime
+  DatasetRunWithItems:
+    extends: DatasetRun
+    properties:
       datasetRunItems: list<DatasetRunItem>
+  Model:
+    docs: Model definition used for transforming usage into USD cost and/or tokenization.
+    properties:
+      id: string
+      modelName:
+        docs: "Name of the model definition. If multiple with the same name exist, they are applied in the following order: (1) custom over built-in, (2) newest according to startTime where model.startTime<observation.startTime"
+        type: string
+      matchPattern:
+        docs: "Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`"
+        type: string
+      startDate:
+        docs: Apply only to generations which are newer than this ISO date.
+        type: optional<date>
+      unit:
+        docs: Unit used by this model.
+        type: ModelUsageUnit
+      inputPrice:
+        docs: Price (USD) per input unit
+        type: optional<double>
+      outputPrice:
+        docs: Price (USD) per output unit
+        type: optional<double>
+      totalPrice:
+        docs: Price (USD) per total unit. Cannot be set if input or output price is set.
+        type: optional<double>
+      tokenizerId:
+        docs: Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
+        type: optional<string>
+      tokenizerConfig:
+        docs: Optional. Configuration for the selected tokenizer. Needs to be JSON. See docs for more details.
+        type: optional<unknown>
+      isLangfuseManaged:
+        type: boolean
 
   # Utilities
   ModelUsageUnit:
@@ -212,6 +314,11 @@ types:
       - ANNOTATION
       - API
       - EVAL
+  ScoreDataType:
+    enum:
+      - NUMERIC
+      - BOOLEAN
+      - CATEGORICAL
 
 errors:
   Error:

fern/apis/server/definition/dataset-items.yml

@@ -1,6 +1,7 @@
 # yaml-language-server: $schema=https://raw.githubusercontent.com/fern-api/fern/main/fern.schema.json
 imports:
   commons: ./commons.yml
+  pagination: ./utils/pagination.yml
 service:
   auth: true
   base-path: /api/public
@@ -19,6 +20,23 @@ service:
         id:
           type: string
       response: commons.DatasetItem
+    list:
+      docs: Get dataset items
+      method: GET
+      path: /dataset-items
+      request:
+        name: GetDatasetItemsRequest
+        query-parameters:
+          datasetName: optional<string>
+          sourceTraceId: optional<string>
+          sourceObservationId: optional<string>
+          page:
+            type: optional<integer>
+            docs: page number, starts at 1
+          limit:
+            type: optional<integer>
+            docs: limit of items per page
+      response: PaginatedDatasetItems
 
 types:
   CreateDatasetItemRequest:
@@ -31,7 +49,11 @@ types:
       sourceObservationId: optional<string>
       id:
         type: optional<string>
-        docs: Dataset items are upserted on their id
+        docs: Dataset items are upserted on their id. Id needs to be globally unique and cannot be reused across datasets.
       status:
         type: optional<commons.DatasetStatus>
         docs: Defaults to ACTIVE for newly created items
+  PaginatedDatasetItems:
+    properties:
+      data: list<commons.DatasetItem>
+      meta: pagination.MetaResponse

fern/apis/server/definition/datasets.yml

@@ -9,42 +9,66 @@ service:
     list:
       method: GET
       docs: Get all datasets
-      path: /datasets
+      path: /v2/datasets
       request:
         name: GetDatasetsRequest
         query-parameters:
-          page: optional<integer>
-          limit: optional<integer>
+          page:
+            type: optional<integer>
+            docs: page number, starts at 1
+          limit:
+            type: optional<integer>
+            docs: limit of items per page
       response: PaginatedDatasets
     get:
       method: GET
-      docs: Get a dataset and its items
-      path: /datasets/{datasetName}
+      docs: Get a dataset
+      path: /v2/datasets/{datasetName}
       path-parameters:
         datasetName: string
       response: commons.Dataset
     create:
       method: POST
       docs: Create a dataset
-      path: /datasets
+      path: /v2/datasets
       request: CreateDatasetRequest
       response: commons.Dataset
-    getRuns:
+    getRun:
       method: GET
       docs: Get a dataset run and its items
       path: /datasets/{datasetName}/runs/{runName}
       path-parameters:
         datasetName: string
         runName: string
-      response: commons.DatasetRun
+      response: commons.DatasetRunWithItems
+    getRuns:
+      method: GET
+      docs: Get dataset runs
+      path: /datasets/{datasetName}/runs
+      path-parameters:
+        datasetName: string
+      request:
+        name: GetDatasetRunsRequest
+        query-parameters:
+          page:
+            type: optional<integer>
+            docs: page number, starts at 1
+          limit:
+            type: optional<integer>
+            docs: limit of items per page
+      response: PaginatedDatasetRuns
 
 types:
   PaginatedDatasets:
     properties:
-      data: list<commons.DatasetWithReferences>
+      data: list<commons.Dataset>
       meta: pagination.MetaResponse
   CreateDatasetRequest:
     properties:
       name: string
       description: optional<string>
       metadata: optional<unknown>
+  PaginatedDatasetRuns:
+    properties:
+      data: list<commons.DatasetRun>
+      meta: pagination.MetaResponse

fern/apis/server/definition/ingestion.yml

@@ -7,28 +7,57 @@ service:
   base-path: /api/public
   endpoints:
     batch:
-      docs: Batched ingestion for Langfuse Tracing
+      docs: |
+        Batched ingestion for Langfuse Tracing. If you want to use tracing via the API, such as to build your own Langfuse client implementation, this is the only API route you need to implement.
+
+        Notes:
+        - Batch sizes are limited to 3.5 MB in total. You need to adjust the number of events per batch accordingly.
+        - The API does not return a 4xx status code for input errors. Instead, it responds with a 207 status code, which includes a list of the encountered errors.
       method: POST
       path: /ingestion
       request:
         name: IngestionRequest
         body:
           properties:
-            batch: list<IngestionEvent>
-      response: IngestionResponse # will be reportet as 200 response, but endpoint returns 207
+            batch:
+              type: list<IngestionEvent>
+              docs: "Batch of tracing events to be ingested. Discriminated by attribute `type`."
+            metadata:
+              type: optional<unknown>
+              docs: Optional. Metadata field used by the Langfuse SDKs for debugging.
+      response:
+        type: IngestionResponse
+        status-code: 207
 
 types:
   IngestionEvent:
     discriminant: "type"
     union:
-      trace-create: TraceEvent
-      score-create: ScoreEvent
-      event-create: CreateEventEvent
-      generation-create: CreateGenerationEvent
-      generation-update: UpdateGenerationEvent
-      span-create: CreateSpanEvent
-      span-update: UpdateSpanEvent
-      sdk-log: SDKLogEvent
+      trace-create:
+        type: TraceEvent
+        docs: Creates a new trace. Upserts on id for updates if trace with id exists.
+      score-create:
+        type: ScoreEvent
+        docs: Creates a new score. Upserts on id for updates if score with id exists.
+      span-create:
+        type: CreateSpanEvent
+        docs: Creates a new span.
+      span-update:
+        type: UpdateSpanEvent
+        docs: Updates span based on id.
+      generation-create:
+        type: CreateGenerationEvent
+        docs: Creates a new generation.
+      generation-update:
+        type: UpdateGenerationEvent
+        docs: Updates a generation based on id.
+      event-create:
+        type: CreateEventEvent
+        docs: Creates an event.
+
+      sdk-log:
+        type: SDKLogEvent
+        docs: Langfuse SDKs only, used for debugging purposes.
 
       # both are legacy
       observation-create:
@@ -156,15 +185,66 @@ types:
       id: optional<string>
       traceId: string
       name: string
-      value: double
+      value:
+        type: commons.CreateScoreValue
+        docs: The value of the score. Must be passed as string for categorical scores, and numeric for boolean and numeric scores. Boolean score values must equal either 1 or 0 (true or false)
       observationId: optional<string>
       comment: optional<string>
+      dataType:
+        type: optional<commons.ScoreDataType>
+        docs: When set, must match the score value's type. If not set, will be inferred from the score value or config
+      configId:
+        type: optional<string>
+        docs: Reference a score config on a score. When set, the score name must equal the config name and scores must comply with the config's range and data type. For categorical scores, the value must map to a config category. Numeric scores might be constrained by the score config's max and min values
+    examples:
+      - value:
+          name: "novelty"
+          value: 0.9
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "consistency"
+          value: 1.2
+          dataType: "NUMERIC"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "accuracy"
+          value: 0.9
+          dataType: "NUMERIC"
+          configId: "9203-4567-89ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "toxicity"
+          value: "not toxic"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "correctness"
+          value: "partially correct"
+          dataType: "CATEGORICAL"
+          configId: "1234-5678-90ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "hallucination"
+          value: 0
+          dataType: "BOOLEAN"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "helpfulness"
+          value: 1
+          dataType: "BOOLEAN"
+          configId: "1234-5678-90ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
 
   BaseEvent:
     properties:
-      id: string
-      timestamp: string
-      metadata: unknown
+      id:
+        type: string
+        docs: UUID v4 that identifies the event
+      timestamp:
+        type: string
+        docs: "Datetime (ISO 8601) of event creation in client. Should be as close to actual event creation in client as possible, this timestamp will be used for ordering of events in future release. Resolution: milliseconds (required), microseconds (optimal)."
+      metadata:
+        type: optional<unknown>
+        docs: Optional. Metadata field used by the Langfuse SDKs for debugging.
 
   TraceEvent:
     extends: BaseEvent

fern/apis/server/definition/metrics.yml

@@ -48,12 +48,25 @@ types:
       date: date
       countTraces: integer
       countObservations: integer
-      totalCost: double
+      totalCost:
+        type: double
+        docs: Total model cost in USD
       usage: list<UsageByModel>
   UsageByModel:
     docs: Daily usage of a given model. Usage corresponds to the unit set for the specific model (e.g. tokens).
     properties:
       model: optional<string>
-      inputUsage: integer
-      outputUsage: integer
-      totalUsage: integer
+      inputUsage:
+        type: integer
+        docs: Total number of generation input units (e.g. tokens)
+      outputUsage:
+        type: integer
+        docs: Total number of generation output units (e.g. tokens)
+      totalUsage:
+        type: integer
+        docs: Total number of generation total units (e.g. tokens)
+      countTraces: integer
+      countObservations: integer
+      totalCost:
+        type: double
+        docs: Total model cost in USD

fern/apis/server/definition/models.yml

@@ -0,0 +1,76 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/fern-api/fern/main/fern.schema.json
+imports:
+  commons: ./commons.yml
+  pagination: ./utils/pagination.yml
+service:
+  auth: true
+  base-path: /api/public
+  endpoints:
+    create:
+      method: POST
+      docs: Create a model
+      path: /models
+      request: CreateModelRequest
+      response: commons.Model
+    list:
+      method: GET
+      docs: Get all models
+      path: /models
+      request:
+        name: GetModelsRequest
+        query-parameters:
+          page:
+            type: optional<integer>
+            docs: page number, starts at 1
+          limit:
+            type: optional<integer>
+            docs: limit of items per page
+      response: PaginatedModels
+    get:
+      method: GET
+      docs: Get a model
+      path: /models/{id}
+      path-parameters:
+        id: string
+      response: commons.Model
+    delete:
+      method: DELETE
+      docs: Delete a model. Cannot delete models managed by Langfuse. You can create your own definition with the same modelName to override the definition though.
+      path: /models/{id}
+      path-parameters:
+        id: string
+
+types:
+  PaginatedModels:
+    properties:
+      data: list<commons.Model>
+      meta: pagination.MetaResponse
+  CreateModelRequest:
+    properties:
+      modelName:
+        docs: "Name of the model definition. If multiple with the same name exist, they are applied in the following order: (1) custom over built-in, (2) newest according to startTime where model.startTime<observation.startTime"
+        type: string
+      matchPattern:
+        docs: "Regex pattern which matches this model definition to generation.model. Useful in case of fine-tuned models. If you want to exact match, use `(?i)^modelname$`"
+        type: string
+      startDate:
+        docs: Apply only to generations which are newer than this ISO date.
+        type: optional<date>
+      unit:
+        docs: Unit used by this model.
+        type: commons.ModelUsageUnit
+      inputPrice:
+        docs: Price (USD) per input unit
+        type: optional<double>
+      outputPrice:
+        docs: Price (USD) per output unit
+        type: optional<double>
+      totalPrice:
+        docs: Price (USD) per total units. Cannot be set if input or output price is set.
+        type: optional<double>
+      tokenizerId:
+        docs: Optional. Tokenizer to be applied to observations which match to this model. See docs for more details.
+        type: optional<string>
+      tokenizerConfig:
+        docs: Optional. Configuration for the selected tokenizer. Needs to be JSON. See docs for more details.
+        type: optional<unknown>

fern/apis/server/definition/observations.yml

@@ -35,7 +35,10 @@ service:
           parentObservationId: optional<string>
           fromStartTime:
             type: optional<datetime>
-            docs: Retrieve only observations with a start_time greater than this datetime (ISO 8601).
+            docs: Retrieve only observations with a start_time or or after this datetime (ISO 8601).
+          toStartTime:
+            type: optional<datetime>
+            docs: Retrieve only observations with a start_time before this datetime (ISO 8601).
       response: ObservationsViews
 
 types:

fern/apis/server/definition/prompts.yml

@@ -35,8 +35,12 @@ service:
           name: optional<string>
           label: optional<string>
           tag: optional<string>
-          page: optional<integer>
-          limit: optional<integer>
+          page:
+            type: optional<integer>
+            docs: page number, starts at 1
+          limit:
+            type: optional<integer>
+            docs: limit of items per page
 
       response: PromptMetaListResponse
 
@@ -70,14 +74,24 @@ types:
       name: string
       prompt: list<ChatMessage>
       config: optional<unknown>
-      labels: optional<list<string>>
+      labels:
+        type: optional<list<string>>
+        docs: List of deployment labels of this prompt version.
+      tags:
+        type: optional<list<string>>
+        docs: List of tags to apply to all versions of this prompt.
 
   CreateTextPromptRequest:
     properties:
       name: string
       prompt: string
       config: optional<unknown>
-      labels: optional<list<string>>
+      labels:
+        type: optional<list<string>>
+        docs: List of deployment labels of this prompt version.
+      tags:
+        type: optional<list<string>>
+        docs: List of tags to apply to all versions of this prompt.
 
   Prompt:
     union:
@@ -89,7 +103,12 @@ types:
       name: string
       version: integer
       config: unknown
-      labels: list<string>
+      labels:
+        type: list<string>
+        docs: List of deployment labels of this prompt version.
+      tags:
+        type: list<string>
+        docs: List of tags. Used to filter via UI and API. The same across versions of a prompt.
 
   ChatMessage:
     properties:

fern/apis/server/definition/score-configs.yml

@@ -0,0 +1,58 @@
+# yaml-language-server: $schema=https://raw.githubusercontent.com/fern-api/fern/main/fern.schema.json
+imports:
+  pagination: ./utils/pagination.yml
+  commons: ./commons.yml
+service:
+  auth: true
+  base-path: /api/public
+  endpoints:
+    create:
+      docs: Create a score configuration (config). Score configs are used to define the structure of scores
+      method: POST
+      path: /score-configs
+      request: CreateScoreConfigRequest
+      response: commons.ScoreConfig
+    get:
+      docs: Get all score configs
+      method: GET
+      path: /score-configs
+      request:
+        name: GetScoreConfigsRequest
+        query-parameters:
+          page:
+            type: optional<integer>
+            docs: Page number, starts at 1.
+          limit:
+            type: optional<integer>
+            docs: Limit of items per page. If you encounter api issues due to too large page sizes, try to reduce the limit
+      response: ScoreConfigs
+    get-by-id:
+      docs: Get a score config
+      method: GET
+      path: /score-configs/{configId}
+      path-parameters:
+        configId:
+          type: string
+          docs: The unique langfuse identifier of a score config
+      response: commons.ScoreConfig
+types:
+  ScoreConfigs:
+    properties:
+      data: list<commons.ScoreConfig>
+      meta: pagination.MetaResponse
+  CreateScoreConfigRequest:
+    properties:
+      name: string
+      dataType: commons.ScoreDataType
+      categories:
+        type: optional<list<commons.ConfigCategory>>
+        docs: Configure custom categories for categorical scores. Pass a list of objects with `label` and `value` properties. Categories are autogenerated for boolean configs and cannot be passed
+      minValue:
+        type: optional<double>
+        docs: Configure a minimum value for numerical scores. If not set, the minimum value defaults to -∞
+      maxValue:
+        type: optional<double>
+        docs: Configure a maximum value for numerical scores. If not set, the maximum value defaults to +∞
+      description:
+        type: optional<string>
+        docs: Description is shown across the Langfuse UI and can be used to e.g. explain the config categories in detail, why a numeric range was set, or provide additional context on config name or usage

fern/apis/server/definition/score.yml

@@ -11,7 +11,7 @@ service:
       method: POST
       path: /scores
       request: CreateScoreRequest
-      response: commons.Score
+      response: CreateScoreResponse
     get:
       docs: Get a list of scores
       method: GET
@@ -25,13 +25,20 @@ service:
           limit:
             type: optional<integer>
             docs: Limit of items per page. If you encounter api issues due to too large page sizes, try to reduce the limit.
-          userId: optional<string>
-          name: optional<string>
+          userId:
+            type: optional<string>
+            docs: Retrieve only scores with this userId associated to the trace.
+          name:
+            type: optional<string>
+            docs: Retrieve only scores with this name.
           fromTimestamp:
             type: optional<datetime>
-            docs: Retrieve only scores newer than this datetime (ISO 8601).
+            docs: Optional filter to only include scores created on or after a certain datetime (ISO 8601)
+          toTimestamp:
+            type: optional<datetime>
+            docs: Optional filter to only include scores created before a certain datetime (ISO 8601)
           source:
-            type: commons.ScoreSource
+            type: optional<commons.ScoreSource>
             docs: Retrieve only scores from a specific source.
           operator:
             type: optional<string>
@@ -39,6 +46,15 @@ service:
           value:
             type: optional<double>
             docs: Retrieve only scores with <operator> value.
+          scoreIds:
+            type: optional<string>
+            docs: Comma-separated list of score IDs to limit the results to.
+          configId:
+            type: optional<string>
+            docs: Retrieve only scores with a specific configId.
+          dataType:
+            type: optional<commons.ScoreDataType>
+            docs: Retrieve only scores with a specific dataType.
       response: Scores
     get-by-id:
       docs: Get a score
@@ -63,9 +79,59 @@ types:
       id: optional<string>
       traceId: string
       name: string
-      value: double
+      value:
+        type: commons.CreateScoreValue
+        docs: The value of the score. Must be passed as string for categorical scores, and numeric for boolean and numeric scores. Boolean score values must equal either 1 or 0 (true or false)
       observationId: optional<string>
       comment: optional<string>
+      dataType:
+        type: optional<commons.ScoreDataType>
+        docs: The data type of the score. When passing a configId this field is inferred. Otherwise, this field must be passed or will default to numeric.
+      configId:
+        type: optional<string>
+        docs: Reference a score config on a score. The unique langfuse identifier of a score config. When passing this field, the dataType and stringValue fields are automatically populated.
+    examples:
+      - value:
+          name: "novelty"
+          value: 0.9
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "consistency"
+          value: 1.2
+          dataType: "NUMERIC"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "accuracy"
+          value: 0.9
+          dataType: "NUMERIC"
+          configId: "9203-4567-89ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "toxicity"
+          value: "not toxic"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "correctness"
+          value: "partially correct"
+          dataType: "CATEGORICAL"
+          configId: "1234-5678-90ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "hallucination"
+          value: 0
+          dataType: "BOOLEAN"
+          traceId: "cdef-1234-5678-90ab"
+      - value:
+          name: "helpfulness"
+          value: 1
+          dataType: "BOOLEAN"
+          configId: "1234-5678-90ab-cdef"
+          traceId: "cdef-1234-5678-90ab"
+  CreateScoreResponse:
+    properties:
+      id:
+        type: string
+        docs: The id of the created object in Langfuse
   Scores:
     properties:
       data: list<commons.Score>

fern/apis/server/definition/sessions.yml

@@ -6,8 +6,28 @@ service:
   auth: true
   base-path: /api/public
   endpoints:
+    list:
+      docs: Get sessions.
+      method: GET
+      path: /sessions
+      request:
+        name: GetSessionsRequest
+        query-parameters:
+          page:
+            type: optional<integer>
+            docs: Page number, starts at 1
+          limit:
+            type: optional<integer>
+            docs: Limit of items per page. If you encounter api issues due to too large page sizes, try to reduce the limit.
+          fromTimestamp:
+            type: optional<datetime>
+            docs: Optional filter to only include sessions created on or after a certain datetime (ISO 8601)
+          toTimestamp:
+            type: optional<datetime>
+            docs: Optional filter to only include sessions created before a certain datetime (ISO 8601)
+        response: PaginatedSessions
     get:
-      docs: Get a session
+      docs: Get a session. Please note that `traces` on this endpoint are not paginated, if you plan to fetch large sessions, consider `GET /api/public/traces?sessionId=<sessionId>`
       method: GET
       path: /sessions/{sessionId}
       path-parameters:
@@ -15,3 +35,8 @@ service:
           type: string
           docs: The unique id of a session
       response: commons.SessionWithTraces
+types:
+  PaginatedSessions:
+    properties:
+      data: list<commons.Session>
+      meta: pagination.MetaResponse

fern/apis/server/definition/trace.yml

@@ -30,9 +30,13 @@ service:
             docs: Limit of items per page. If you encounter api issues due to too large page sizes, try to reduce the limit.
           userId: optional<string>
           name: optional<string>
+          sessionId: optional<string>
           fromTimestamp:
             type: optional<datetime>
-            docs: Retrieve only traces newer than this datetime (ISO 8601).
+            docs: Optional filter to only include traces with a trace.timestamp on or after a certain datetime (ISO 8601)
+          toTimestamp:
+            type: optional<datetime>
+            docs: Optional filter to only include traces with a trace.timestamp before a certain datetime (ISO 8601)
           orderBy:
             type: optional<string>
             docs: "Format of the string [field].[asc/desc]. Fields: id, timestamp, name, userId, release, version, public, bookmarked, sessionId. Example: timestamp.asc"

fern/apis/server/definition/utils/pagination.yml

@@ -1,4 +1,13 @@
 types:
+  # to be used as copy/paste template for query strings
+  # PaginationRequest:
+  #   properties:
+  #     page:
+  #       type: optional<integer>
+  #       docs: page number, starts at 1
+  #     limit:
+  #       type: optional<integer>
+  #       docs: limit of items per page
   MetaResponse:
     properties:
       page:

fern/fern.config.json

@@ -1,4 +1,4 @@
 {
   "organization": "finto",
-  "version": "0.21.0"
+  "version": "0.30.3"
 }

package.json

@@ -1,6 +1,6 @@
 {
   "name": "langfuse",
-  "version": "2.47.2",
+  "version": "2.61.0",
   "author": "engineering@langfuse.com",
   "license": "MIT",
   "private": true,
@@ -17,26 +17,36 @@
     "db:seed": "turbo run db:seed",
     "db:seed:examples": "turbo run db:seed:examples",
     "nuke": "bash ./scripts/nuke.sh",
-    "dx": "pnpm i && pnpm --filter=shared run db:reset && pnpm --filter=shared run db:seed:examples && pnpm run dev",
+    "dx": "pnpm i && pnpm run infra:dev:up && pnpm --filter=shared run db:reset && pnpm --filter=shared run db:seed:examples && pnpm run dev",
+    "dx-f": "pnpm i && pnpm run infra:dev:up && pnpm --filter=shared run db:reset -f && pnpm --filter=shared run db:seed:examples && pnpm run dev",
     "build": "turbo run build",
     "start": "turbo run start",
     "dev": "turbo run dev",
     "lint": "turbo run lint",
     "test": "turbo run test",
     "models:migrate": "turbo run models:migrate",
-    "release": "dotenv -e ../.env -- release-it"
+    "release": "dotenv -e ../.env -- release-it",
+    "prepare": "husky"
   },
   "devDependencies": {
     "@release-it/bumper": "^6.0.1",
+    "braces": "3.0.3",
     "dotenv-cli": "^7.4.2",
+    "husky": "^9.0.11",
     "prettier": "^3.2.5",
-    "release-it": "^17.2.1",
+    "release-it": "^17.3.0",
     "turbo": "^1.13.3"
   },
   "release-it": {
     "git": {
       "commitMessage": "chore: release v${version}",
-      "tagName": "v${version}"
+      "tagName": "v${version}",
+      "commitArgs": [
+        "--no-verify"
+      ],
+      "pushArgs": [
+        "--no-verify"
+      ]
     },
     "plugins": {
       "@release-it/bumper": {

packages/config-eslint/package.json

@@ -9,7 +9,7 @@
   ],
   "devDependencies": {
     "@typescript-eslint/eslint-plugin": "^7.1.0",
-    "@typescript-eslint/parser": "^7.7.0",
+    "@typescript-eslint/parser": "^7.12.0",
     "@vercel/style-guide": "^6.0.0",
     "eslint-config-next": "^14.2.3",
     "eslint-config-prettier": "^9.1.0",

packages/shared/package.json

@@ -17,9 +17,9 @@
       "import": "./dist/src/db.js",
       "require": "./dist/src/db.js"
     },
-    "./src/server/auth": {
-      "import": "./dist/src/server/auth.js",
-      "require": "./dist/src/server/auth.js"
+    "./src/server": {
+      "import": "./dist/src/server/index.js",
+      "require": "./dist/src/server/index.js"
     },
     "./encryption": {
       "import": "./dist/src/encryption/index.js",
@@ -44,32 +44,41 @@
     "seed": "ts-node -r tsconfig-paths/register -r dotenv/config --compiler-options {\"module\":\"CommonJS\"} prisma/seed.ts"
   },
   "dependencies": {
+    "@aws-sdk/client-s3": "^3.550.0",
+    "@aws-sdk/lib-storage": "^3.568.0",
+    "@aws-sdk/s3-request-presigner": "^3.554.0",
     "@langchain/anthropic": "^0.1.21",
     "@langchain/core": "^0.1.61",
     "@langchain/openai": "^0.0.33",
     "@prisma/client": "^5.13.0",
+    "@react-email/components": "^0.0.19",
+    "@react-email/render": "^0.0.15",
     "@types/bcryptjs": "^2.4.6",
     "bcryptjs": "^2.4.3",
     "kysely": "^0.27.3",
-    "langchain": "^0.1.36",
+    "langchain": "^0.2.6",
+    "lodash": "^4.17.21",
+    "next-auth": "^4.24.7",
+    "nodemailer": "^6.9.13",
     "prisma-extension-kysely": "^2.1.0",
-    "zod": "^3.22.4",
+    "zod": "^3.23.8",
     "zod-to-json-schema": "^3.22.5"
   },
   "devDependencies": {
     "@repo/eslint-config": "*",
     "@repo/typescript-config": "*",
-    "@types/lodash": "^4.17.4",
+    "@types/lodash": "^4.17.5",
     "@types/node": "^20.11.29",
+    "@types/nodemailer": "^6.4.15",
     "@types/pg": "^8.11.6",
     "@types/uuid": "^9.0.8",
-    "@typescript-eslint/parser": "^7.7.0",
+    "@typescript-eslint/parser": "^7.12.0",
     "eslint": "^8.57.0",
     "eslint-config-prettier": "^9.1.0",
     "eslint-config-standard": "^17.1.0",
     "eslint-plugin-prettier": "^5.1.3",
     "kysely-codegen": "^0.11.0",
-    "nodemon": "^3.0.3",
+    "nodemon": "^3.1.3",
     "prettier": "^3.2.5",
     "prisma": "^5.13.0",
     "prisma-erd-generator": "^1.11.2",
@@ -78,5 +87,9 @@
     "tsc-watch": "^6.2.0",
     "typescript": "^5.4.5",
     "vitest": "^1.5.3"
+  },
+  "peerDependencies": {
+    "@types/react": "^18.2.79",
+    "react": "^18.0.0"
   }
 }

packages/shared/prisma/generated/types.ts

@@ -32,20 +32,10 @@ export const ScoreSource = {
 export type ScoreSource = (typeof ScoreSource)[keyof typeof ScoreSource];
 export const ScoreDataType = {
     CATEGORICAL: "CATEGORICAL",
-    NUMERIC: "NUMERIC"
+    NUMERIC: "NUMERIC",
+    BOOLEAN: "BOOLEAN"
 } as const;
 export type ScoreDataType = (typeof ScoreDataType)[keyof typeof ScoreDataType];
-export const PricingUnit = {
-    PER_1000_TOKENS: "PER_1000_TOKENS",
-    PER_1000_CHARS: "PER_1000_CHARS"
-} as const;
-export type PricingUnit = (typeof PricingUnit)[keyof typeof PricingUnit];
-export const TokenType = {
-    PROMPT: "PROMPT",
-    COMPLETION: "COMPLETION",
-    TOTAL: "TOTAL"
-} as const;
-export type TokenType = (typeof TokenType)[keyof typeof TokenType];
 export const DatasetStatus = {
     ACTIVE: "ACTIVE",
     ARCHIVED: "ARCHIVED"
@@ -108,6 +98,21 @@ export type AuditLog = {
     before: string | null;
     after: string | null;
 };
+export type BatchExport = {
+    id: string;
+    created_at: Generated<Timestamp>;
+    updated_at: Generated<Timestamp>;
+    project_id: string;
+    user_id: string;
+    finished_at: Timestamp | null;
+    expires_at: Timestamp | null;
+    name: string;
+    status: string;
+    query: unknown;
+    format: string;
+    url: string | null;
+    log: string | null;
+};
 export type CronJobs = {
     name: string;
     last_run: Timestamp | null;
@@ -162,6 +167,7 @@ export type EvalTemplate = {
     version: number;
     prompt: string;
     model: string;
+    provider: string;
     model_params: unknown;
     vars: Generated<string[]>;
     output_schema: unknown;
@@ -209,8 +215,12 @@ export type LlmApiKeys = {
     created_at: Generated<Timestamp>;
     updated_at: Generated<Timestamp>;
     provider: string;
+    adapter: string;
     display_secret_key: string;
     secret_key: string;
+    base_url: string | null;
+    custom_models: Generated<string[]>;
+    with_default_models: Generated<boolean>;
     project_id: string;
 };
 export type MembershipInvitation = {
@@ -254,6 +264,7 @@ export type Observation = {
     updated_at: Generated<Timestamp>;
     model: string | null;
     internal_model: string | null;
+    internal_model_id: string | null;
     modelParameters: unknown | null;
     input: unknown | null;
     output: unknown | null;
@@ -264,6 +275,9 @@ export type Observation = {
     input_cost: string | null;
     output_cost: string | null;
     total_cost: string | null;
+    calculated_input_cost: string | null;
+    calculated_output_cost: string | null;
+    calculated_total_cost: string | null;
     completion_start_time: Timestamp | null;
     prompt_id: string | null;
 };
@@ -272,7 +286,7 @@ export type ObservationView = {
     trace_id: string | null;
     project_id: string;
     type: ObservationType;
-    start_time: Generated<Timestamp>;
+    start_time: Timestamp;
     end_time: Timestamp | null;
     name: string | null;
     metadata: unknown | null;
@@ -280,7 +294,8 @@ export type ObservationView = {
     level: Generated<ObservationLevel>;
     status_message: string | null;
     version: string | null;
-    created_at: Generated<Timestamp>;
+    created_at: Timestamp;
+    updated_at: Timestamp;
     model: string | null;
     modelParameters: unknown | null;
     input: unknown | null;
@@ -309,14 +324,6 @@ export type PosthogIntegration = {
     enabled: boolean;
     created_at: Generated<Timestamp>;
 };
-export type Pricing = {
-    id: string;
-    model_name: string;
-    pricing_unit: Generated<PricingUnit>;
-    price: string;
-    currency: Generated<string>;
-    token_type: TokenType;
-};
 export type Project = {
     id: string;
     created_at: Generated<Timestamp>;
@@ -351,7 +358,7 @@ export type Score = {
     timestamp: Generated<Timestamp>;
     project_id: string;
     name: string;
-    value: number;
+    value: number | null;
     source: ScoreSource;
     author_user_id: string | null;
     comment: string | null;
@@ -432,6 +439,8 @@ export type TraceView = {
     input: unknown | null;
     output: unknown | null;
     session_id: string | null;
+    created_at: Timestamp;
+    updated_at: Timestamp;
     duration: number | null;
 };
 export type User = {
@@ -455,6 +464,7 @@ export type DB = {
     Account: Account;
     api_keys: ApiKey;
     audit_logs: AuditLog;
+    batch_exports: BatchExport;
     cron_jobs: CronJobs;
     dataset_items: DatasetItem;
     dataset_run_items: DatasetRunItems;
@@ -470,7 +480,6 @@ export type DB = {
     observations: Observation;
     observations_view: ObservationView;
     posthog_integrations: PosthogIntegration;
-    pricings: Pricing;
     project_memberships: ProjectMembership;
     projects: Project;
     prompts: Prompt;

packages/shared/prisma/migrations/20240530212419_model_price_anthropic_via_google_vertex/migration.sql

@@ -1,7 +0,0 @@
--- Google Vertex uses @ to separate model name and version
-
-UPDATE "public"."models" SET "match_pattern" = '(?i)^(claude-3-haiku(-|@)?20240307)$' WHERE "id" = 'cltr0w45b000008k1407o9qv1';
-
-UPDATE "public"."models" SET "match_pattern" = '(?i)^(claude-3-opus(-|@)?20240229)$' WHERE "id" = 'cltgy0iuw000008le3vod1hhy';
-
-UPDATE "public"."models" SET "match_pattern" = '(?i)^(claude-3-sonnet(-|@)?20240229)$' WHERE "id" = 'cltgy0pp6000108le56se7bl3';

packages/shared/prisma/migrations/20240603212024_dataset_items_add_index_source_trace_id/migration.sql

@@ -0,0 +1,2 @@
+-- CreateIndex
+CREATE INDEX CONCURRENTLY "dataset_items_source_trace_id_idx" ON "dataset_items" USING HASH ("source_trace_id");

packages/shared/prisma/migrations/20240604133338_scores_add_index_name/migration.sql

@@ -0,0 +1,2 @@
+-- CreateIndex
+CREATE INDEX "scores_project_id_name_idx" ON "scores"("project_id", "name");

packages/shared/prisma/migrations/20240604133339_score_data_type_add_boolean/migration.sql

@@ -0,0 +1,2 @@
+-- AlterEnum
+ALTER TYPE "ScoreDataType" ADD VALUE 'BOOLEAN';

packages/shared/prisma/migrations/20240606093356_drop_unused_pricings_table/migration.sql

@@ -0,0 +1,14 @@
+/*
+  Warnings:
+
+  - You are about to drop the `pricings` table. If the table is not empty, all the data it contains will be lost.
+
+*/
+-- DropTable
+DROP TABLE "pricings";
+
+-- DropEnum
+DROP TYPE "PricingUnit";
+
+-- DropEnum
+DROP TYPE "TokenType";

packages/shared/prisma/migrations/20240606133011_remove_trace_fkey_datasetrunitems/migration.sql

@@ -0,0 +1,3 @@
+-- DropForeignKey
+ALTER TABLE "dataset_run_items" DROP CONSTRAINT "dataset_run_items_trace_id_fkey";
+ALTER TABLE "dataset_run_items" DROP CONSTRAINT "dataset_run_items_observation_id_fkey";

packages/shared/prisma/migrations/20240607090858_pricings_add_latest_gemini_models/migration.sql

@@ -0,0 +1,8 @@
+-- does not include pricing yet, will be added as soon as it is calculated at ingestion time
+
+-- remove model if added by faulty migration, context: https://github.com/langfuse/langfuse/issues/2266
+DELETE FROM models WHERE id in ('clx30djsn0000w9mzebiv41we', 'clx30hkrx0000w9mz7lqi0ial');
+
+INSERT INTO "models" ("id", "model_name", "match_pattern", "unit") VALUES ('clx30djsn0000w9mzebiv41we', 'gemini-1.5-flash', '(?i)^(gemini-1.5-flash)(@[a-zA-Z0-9]+)?$', 'CHARACTERS');
+
+INSERT INTO "models" ("id", "model_name", "match_pattern", "unit") VALUES ('clx30hkrx0000w9mz7lqi0ial', 'gemini-1.5-pro', '(?i)^(gemini-1.5-pro)(@[a-zA-Z0-9]+)?$', 'CHARACTERS');

packages/shared/prisma/migrations/20240607212419_model_price_anthropic_via_google_vertex/migration.sql

@@ -0,0 +1,7 @@
+-- Google Vertex uses @ to separate model name and version
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-haiku(-|@)?20240307)$' WHERE "id" = 'cltr0w45b000008k1407o9qv1';
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-opus(-|@)?20240229)$' WHERE "id" = 'cltgy0iuw000008le3vod1hhy';
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-sonnet(-|@)?20240229)$' WHERE "id" = 'cltgy0pp6000108le56se7bl3';

packages/shared/prisma/migrations/20240611105521_llm_api_keys_custom_endpoints/migration.sql

@@ -0,0 +1,24 @@
+BEGIN;
+
+  ALTER TABLE "llm_api_keys" 
+    ADD COLUMN "base_url" TEXT,
+    ADD COLUMN "adapter" TEXT,
+    ADD COLUMN "custom_models" TEXT[] NOT NULL DEFAULT '{}'::TEXT[],
+    ADD COLUMN "with_default_models" BOOLEAN NOT NULL DEFAULT true;
+
+  UPDATE "llm_api_keys"
+    SET "adapter" = "provider";
+
+  ALTER TABLE "llm_api_keys" 
+    ALTER COLUMN "adapter" SET NOT NULL;
+
+  ALTER TABLE "eval_templates"
+    ADD COLUMN "provider" TEXT;
+  
+  UPDATE "eval_templates"
+    SET "provider" = 'openai';
+  
+  ALTER TABLE "eval_templates"
+    ALTER COLUMN "provider" SET NOT NULL;
+
+COMMIT;

packages/shared/prisma/migrations/20240611113517_backfill_manual_score_configs/migration.sql

@@ -0,0 +1,51 @@
+-- Create manual-score config for any project with manual scores and link config_id to scores
+BEGIN;
+WITH project_configs AS (
+INSERT INTO score_configs (id,
+		project_id,
+		name,
+		data_type,
+		is_archived,
+		min_value,
+		max_value,
+		description)
+SELECT
+	md5(random()::text || clock_timestamp()::text || s.project_id::text)::uuid AS id,
+	s.project_id,
+	'manual-score',
+	'NUMERIC',
+	FALSE,
+	- 1,
+	1,
+	'Langfuse legacy annotation score.'
+FROM ( SELECT DISTINCT
+		project_id
+	FROM
+		scores
+	WHERE
+		name = 'manual-score'
+		AND config_id IS NULL
+		AND source = 'ANNOTATION') s
+WHERE
+	NOT EXISTS (
+		SELECT
+			1
+		FROM
+			score_configs sc
+		WHERE
+			sc.name = 'manual-score'
+			AND sc.project_id = s.project_id)
+	RETURNING
+		id,
+		project_id
+)
+UPDATE
+	scores
+SET
+	config_id = pc.id
+FROM
+	project_configs pc
+WHERE
+	scores.project_id = pc.project_id
+	AND scores.name = 'manual-score';
+COMMIT;

packages/shared/prisma/migrations/20240612101858_add_index_observations_project_id_prompt_id/migration.sql

@@ -0,0 +1,2 @@
+-- CreateIndex
+CREATE INDEX CONCURRENTLY "observations_project_id_prompt_id_idx" ON "observations"("project_id", "prompt_id");

packages/shared/prisma/migrations/20240617094803_observations_remove_prompt_fk_constraint/migration.sql

@@ -0,0 +1,2 @@
+-- DropForeignKey
+ALTER TABLE "observations" DROP CONSTRAINT "observations_prompt_id_fkey";

packages/shared/prisma/migrations/20240618134129_add_batch_exports_table/migration.sql

@@ -0,0 +1,27 @@
+-- CreateTable
+CREATE TABLE "batch_exports" (
+    "id" TEXT NOT NULL,
+    "created_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "updated_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "project_id" TEXT NOT NULL,
+    "user_id" TEXT NOT NULL,
+    "finished_at" TIMESTAMP(3),
+    "expires_at" TIMESTAMP(3),
+    "name" TEXT NOT NULL,
+    "status" TEXT NOT NULL,
+    "query" JSONB NOT NULL,
+    "format" TEXT NOT NULL,
+    "url" TEXT,
+    "log" TEXT,
+
+    CONSTRAINT "batch_exports_pkey" PRIMARY KEY ("id")
+);
+
+-- CreateIndex
+CREATE INDEX "batch_exports_project_id_user_id_idx" ON "batch_exports"("project_id", "user_id");
+
+-- CreateIndex
+CREATE INDEX "batch_exports_status_idx" ON "batch_exports"("status");
+
+-- AddForeignKey
+ALTER TABLE "batch_exports" ADD CONSTRAINT "batch_exports_project_id_fkey" FOREIGN KEY ("project_id") REFERENCES "projects"("id") ON DELETE CASCADE ON UPDATE CASCADE;

packages/shared/prisma/migrations/20240618164950_drop_observations_parent_observation_id_idx/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY IF EXISTS "observations_parent_observation_id_idx";

packages/shared/prisma/migrations/20240618164951_drop_observations_updated_at_idx/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY IF EXISTS "observations_updated_at_idx";

packages/shared/prisma/migrations/20240618164952_drop_scores_updated_at_idx/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY IF EXISTS "scores_updated_at_idx";

packages/shared/prisma/migrations/20240618164953_drop_traces_external_id_idx/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY IF EXISTS "traces_external_id_idx";

packages/shared/prisma/migrations/20240618164954_drop_traces_release_idx/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY IF EXISTS "traces_release_idx";

packages/shared/prisma/migrations/20240618164955_drop_traces_updated_at_idx/migration.sql

@@ -0,0 +1,2 @@
+-- DropIndex
+DROP INDEX CONCURRENTLY IF EXISTS "traces_updated_at_idx";

packages/shared/prisma/migrations/20240618164956_create_traces_project_id_timestamp_idx/migration.sql

@@ -0,0 +1,2 @@
+-- CreateIndex
+CREATE INDEX CONCURRENTLY "traces_project_id_timestamp_idx" ON "traces"("project_id", "timestamp");

packages/shared/prisma/migrations/20240624133412_models_add_anthropic_3_5_sonnet/migration.sql

@@ -0,0 +1,15 @@
+INSERT INTO models (
+  id,
+  project_id,
+  model_name,
+  match_pattern,
+  start_date,
+  input_price,
+  output_price,
+  total_price,
+  unit,
+  tokenizer_id
+)
+VALUES
+  -- add 3.5 sonnet model
+  ('clxt0n0m60000pumz1j5b7zsf', NULL, 'claude-3-5-sonnet-20240620', '(?i)^(claude-3-5-sonnet(-|@)?20240620)$', NULL, 0.000003, 0.000015, NULL, 'TOKENS', 'claude')

packages/shared/prisma/migrations/20240625103957_observations_add_calculated_cost_columns/migration.sql

@@ -0,0 +1,5 @@
+-- AlterTable
+ALTER TABLE "observations" ADD COLUMN     "calculated_input_cost" DECIMAL(65,30),
+ADD COLUMN     "calculated_output_cost" DECIMAL(65,30),
+ADD COLUMN     "calculated_total_cost" DECIMAL(65,30),
+ADD COLUMN     "internal_model_id" TEXT;

packages/shared/prisma/migrations/20240625103958_fix_model_match_gpt4_vision/migration.sql

@@ -0,0 +1,17 @@
+/*
+Previously, the pattern did not allow for model specifications like "1106" or any other 4-digit block. 
+This led to missing models on the generation-update call where the exact model name that was used was provided (including the 4-digit block).
+The new pattern allows for:
+
+- "gpt-4-vision-preview" as set on generation-create
+- "gpt-4-1106-vision-preview" as set on generation-update
+
+*/
+
+UPDATE
+	"models"
+SET
+	"match_pattern" = '(?i)^(gpt-4(-\d{4})?-vision-preview)$'
+WHERE
+	"id" = 'clrkvx5gp000108juaogs54ea'
+	AND "model_name" = 'gpt-4-turbo-vision';

packages/shared/prisma/migrations/20240703214747_models_anthropic_aws_bedrock/migration.sql

@@ -0,0 +1,9 @@
+-- Add AWS Bedrock model names for Anthropic models
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-haiku-20240307|anthropic\.claude-3-haiku-20240307-v1:0|claude-3-haiku@20240307)$' WHERE "id" = 'cltr0w45b000008k1407o9qv1';
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-sonnet-20240229|anthropic\.claude-3-sonnet-20240229-v1:0|claude-3-sonnet@20240229)$' WHERE "id" = 'cltgy0pp6000108le56se7bl3';
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-opus-20240229|anthropic\.claude-3-opus-20240229-v1:0|claude-3-opus@20240229)$' WHERE "id" = 'cltgy0iuw000008le3vod1hhy';
+
+UPDATE "models" SET "match_pattern" = '(?i)^(claude-3-5-sonnet-20240620|anthropic\.claude-3-5-sonnet-20240620-v1:0|claude-3-5-sonnet@20240620)$' WHERE "id" = 'clxt0n0m60000pumz1j5b7zsf';

packages/shared/prisma/migrations/20240704103900_observations_view_read_from_calculated/migration.sql

@@ -0,0 +1,88 @@
+DROP VIEW IF EXISTS "observations_view"; -- Drop view as column was added in 20240528214726_add_cursor_new_columns_observations and update view must have same columns
+CREATE VIEW "observations_view" AS -- Specify the columns that should be returned in the view, as calculated columns are added but exist in the observations table already
+SELECT
+    o.id,
+    o.name,
+    o.start_time,
+    o.end_time,
+    o.parent_observation_id,
+    o.type,
+    o.trace_id,
+    o.metadata,
+    o.model,
+    o."modelParameters",
+    o.input,
+    o.output,
+    o.level,
+    o.status_message,
+    o.completion_start_time,
+    o.completion_tokens,
+    o.prompt_tokens,
+    o.total_tokens,
+    o.version,
+    o.project_id,
+    o.created_at,
+    o.unit,
+    o.prompt_id,
+    o.input_cost,
+    o.output_cost,
+    o.total_cost,
+    o.internal_model,
+    m.id AS "model_id",
+    m.start_date AS "model_start_date",
+    m.input_price,
+    m.output_price,
+    m.total_price,
+    m.tokenizer_config AS "tokenizer_config",
+    CASE 
+        WHEN o.calculated_input_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            o.prompt_tokens::decimal * m.input_price
+        ELSE
+            COALESCE(o.calculated_input_cost, o.input_cost)
+    END AS "calculated_input_cost",
+    CASE 
+        WHEN o.calculated_output_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            o.completion_tokens::decimal * m.output_price
+        ELSE
+            COALESCE(o.calculated_output_cost, o.output_cost)
+    END AS "calculated_output_cost",
+    CASE 
+        WHEN o.calculated_total_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            CASE 
+                WHEN m.total_price IS NOT NULL AND o.total_tokens IS NOT NULL THEN
+                    m.total_price * o.total_tokens
+                ELSE
+                    o.prompt_tokens::decimal * m.input_price + 
+                    o.completion_tokens::decimal * m.output_price
+            END
+        ELSE
+            COALESCE(o.calculated_total_cost, o.total_cost)
+    END AS "calculated_total_cost",
+    CASE WHEN o.end_time IS NULL THEN NULL ELSE (EXTRACT(EPOCH FROM o."end_time") - EXTRACT(EPOCH FROM o."start_time"))::double precision END AS "latency",
+    CASE WHEN o.completion_start_time IS NOT NULL AND o.start_time IS NOT NULL THEN EXTRACT(EPOCH FROM (completion_start_time - start_time))::double precision ELSE NULL END as "time_to_first_token"
+    
+FROM
+    observations o
+LEFT JOIN LATERAL (
+    SELECT
+        models.*
+    FROM
+        models
+    WHERE (models.project_id = o.project_id OR models.project_id IS NULL)
+    AND models.model_name = o.internal_model
+    AND (models.start_date < o.start_time OR models.start_date IS NULL)
+    AND o.unit::TEXT = models.unit
+    ORDER BY
+        models.project_id ASC, -- in postgres, NULLs are sorted last when ordering ASC
+        models.start_date DESC NULLS LAST -- now, NULLs are sorted last when ordering DESC as well
+    LIMIT 1
+) m ON TRUE
+
+
+-- requirements:
+-- 1. The view should return all columns from the observations table
+-- 2. The view should match with only one model for each observation if:
+--     a. The model has the same project_id as the observation, otherwise the model without project_id. 
+--     b. The model has the same model_name as the observation
+--     c. The model has a start_date that is less than the observation start_time, otherwise the model without start_date
+--     d. The model has the same unit as the observation

packages/shared/prisma/migrations/20240704103901_scores_make_value_optional/migration.sql

@@ -0,0 +1,2 @@
+-- AlterTable
+ALTER TABLE "scores" ALTER COLUMN "value" DROP NOT NULL;

packages/shared/prisma/migrations/20240705152639_traces_view_add_created_at_updated_at/migration.sql

@@ -0,0 +1,20 @@
+-- Drop and create to be able to change columns, otherwise new t.* cols cannot be added
+
+DROP VIEW IF EXISTS traces_view;
+CREATE VIEW traces_view AS
+WITH observations_metrics AS (
+    SELECT
+        trace_id,
+        project_id,
+        EXTRACT(EPOCH FROM COALESCE(MAX(o.end_time), MAX(o.start_time))) - EXTRACT(EPOCH FROM MIN(o.start_time))::double precision AS duration
+    FROM
+        observations o
+    GROUP BY
+        project_id, trace_id
+)
+SELECT
+    t.*,
+    o.duration
+FROM
+    traces t
+    LEFT JOIN observations_metrics o ON t.id = o.trace_id and t.project_id = o.project_id

packages/shared/prisma/migrations/20240705154048_observation_view_add_created_at_updated_at/migration.sql

@@ -0,0 +1,89 @@
+DROP VIEW IF EXISTS "observations_view"; -- Drop view as column was added in 20240704103900_observations_view_read_from_calculated and update view must have same columns
+CREATE VIEW "observations_view" AS -- Specify the columns that should be returned in the view, as calculated columns are added but exist in the observations table already
+SELECT
+    o.id,
+    o.name,
+    o.start_time,
+    o.end_time,
+    o.parent_observation_id,
+    o.type,
+    o.trace_id,
+    o.metadata,
+    o.model,
+    o."modelParameters",
+    o.input,
+    o.output,
+    o.level,
+    o.status_message,
+    o.completion_start_time,
+    o.completion_tokens,
+    o.prompt_tokens,
+    o.total_tokens,
+    o.version,
+    o.project_id,
+    o.created_at,
+    o.updated_at,
+    o.unit,
+    o.prompt_id,
+    o.input_cost,
+    o.output_cost,
+    o.total_cost,
+    o.internal_model,
+    m.id AS "model_id",
+    m.start_date AS "model_start_date",
+    m.input_price,
+    m.output_price,
+    m.total_price,
+    m.tokenizer_config AS "tokenizer_config",
+    CASE 
+        WHEN o.calculated_input_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            o.prompt_tokens::decimal * m.input_price
+        ELSE
+            COALESCE(o.calculated_input_cost, o.input_cost)
+    END AS "calculated_input_cost",
+    CASE 
+        WHEN o.calculated_output_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            o.completion_tokens::decimal * m.output_price
+        ELSE
+            COALESCE(o.calculated_output_cost, o.output_cost)
+    END AS "calculated_output_cost",
+    CASE 
+        WHEN o.calculated_total_cost IS NULL AND o.input_cost IS NULL AND o.output_cost IS NULL AND o.total_cost IS NULL THEN
+            CASE 
+                WHEN m.total_price IS NOT NULL AND o.total_tokens IS NOT NULL THEN
+                    m.total_price * o.total_tokens
+                ELSE
+                    o.prompt_tokens::decimal * m.input_price + 
+                    o.completion_tokens::decimal * m.output_price
+            END
+        ELSE
+            COALESCE(o.calculated_total_cost, o.total_cost)
+    END AS "calculated_total_cost",
+    CASE WHEN o.end_time IS NULL THEN NULL ELSE (EXTRACT(EPOCH FROM o."end_time") - EXTRACT(EPOCH FROM o."start_time"))::double precision END AS "latency",
+    CASE WHEN o.completion_start_time IS NOT NULL AND o.start_time IS NOT NULL THEN EXTRACT(EPOCH FROM (completion_start_time - start_time))::double precision ELSE NULL END as "time_to_first_token"
+    
+FROM
+    observations o
+LEFT JOIN LATERAL (
+    SELECT
+        models.*
+    FROM
+        models
+    WHERE (models.project_id = o.project_id OR models.project_id IS NULL)
+    AND models.model_name = o.internal_model
+    AND (models.start_date < o.start_time OR models.start_date IS NULL)
+    AND o.unit::TEXT = models.unit
+    ORDER BY
+        models.project_id ASC, -- in postgres, NULLs are sorted last when ordering ASC
+        models.start_date DESC NULLS LAST -- now, NULLs are sorted last when ordering DESC as well
+    LIMIT 1
+) m ON TRUE
+
+
+-- requirements:
+-- 1. The view should return all columns from the observations table
+-- 2. The view should match with only one model for each observation if:
+--     a. The model has the same project_id as the observation, otherwise the model without project_id. 
+--     b. The model has the same model_name as the observation
+--     c. The model has a start_date that is less than the observation start_time, otherwise the model without start_date
+--     d. The model has the same unit as the observation

packages/shared/prisma/migrations/20240710114043_score_configs_drop_empty_categories_array_for_numeric_scores/migration.sql

@@ -0,0 +1,6 @@
+-- Migration script to update score_configs entries
+-- Set categories to NULL where data_type is 'NUMERIC' and categories is an empty array
+
+UPDATE score_configs
+SET categories = NULL
+WHERE data_type = 'NUMERIC' AND categories IS NOT NULL;

packages/shared/prisma/schema.prisma

@@ -115,6 +115,7 @@ model Project {
     PosthogIntegration PosthogIntegration[]
     Score              Score[]
     scoreConfig        ScoreConfig[]
+    BatchExport        BatchExport[]
 
     @@map("projects")
 }
@@ -144,9 +145,13 @@ model LlmApiKeys {
     createdAt DateTime @default(now()) @map("created_at")
     updatedAt DateTime @default(now()) @updatedAt @map("updated_at")
 
-    provider         String
-    displaySecretKey String @map("display_secret_key")
-    secretKey        String @map("secret_key")
+    provider          String
+    adapter           String // This controls the interface that is used to connect with the LLM, e.g. 'openai' or 'anthropic'
+    displaySecretKey  String   @map("display_secret_key")
+    secretKey         String   @map("secret_key")
+    baseURL           String?  @map("base_url")
+    customModels      String[] @default([]) @map("custom_models")
+    withDefaultModels Boolean  @default(true) @map("with_default_models")
 
     projectId String  @map("project_id")
     project   Project @relation(fields: [projectId], references: [id], onDelete: Cascade)
@@ -233,19 +238,16 @@ model Trace {
     createdAt  DateTime      @default(now()) @map("created_at")
     updatedAt  DateTime      @default(now()) @updatedAt @map("updated_at")
 
-    DatasetRunItems DatasetRunItems[]
-    DatasetItem     DatasetItem[]
-    JobExecution    JobExecution[]
+    DatasetItem  DatasetItem[]
+    JobExecution JobExecution[]
 
     @@index([projectId])
+    @@index([projectId, timestamp])
     @@index([sessionId])
     @@index([name])
     @@index([userId])
     @@index([id, userId])
-    @@index([externalId])
     @@index(timestamp)
-    @@index(release)
-    @@index(updatedAt)
     @@index(createdAt)
     @@index([tags(ops: ArrayOps)], type: Gin)
     @@map("traces")
@@ -270,6 +272,8 @@ view TraceView {
     input      Json?
     output     Json?
     sessionId  String?  @map("session_id")
+    createdAt  DateTime @map("created_at")
+    updatedAt  DateTime @map("updated_at")
 
     // calculated fields
     duration Float? @map("duration") // can be null if no observations in trace
@@ -297,25 +301,33 @@ model Observation {
     updatedAt           DateTime         @default(now()) @updatedAt @map("updated_at")
 
     // GENERATION ONLY
-    model               String?
-    internalModel       String?           @map("internal_model")
-    modelParameters     Json?
-    input               Json?
-    output              Json?
-    promptTokens        Int               @default(0) @map("prompt_tokens")
-    completionTokens    Int               @default(0) @map("completion_tokens")
-    totalTokens         Int               @default(0) @map("total_tokens")
-    unit                String?
-    inputCost           Decimal?          @map("input_cost")
-    outputCost          Decimal?          @map("output_cost")
-    totalCost           Decimal?          @map("total_cost")
-    completionStartTime DateTime?         @map("completion_start_time")
-    project             Project           @relation(fields: [projectId], references: [id], onDelete: Cascade)
-    derivedDatasetItems DatasetItem[]
-    datasetRunItems     DatasetRunItems[]
+    model           String? // user-provided model attribute
+    internalModel   String? @map("internal_model") // matched model.name that is matched at ingestion time, to be deprecated
+    internalModelId String? @map("internal_model_id") // matched model.id that is matched at ingestion time
 
-    promptId String? @map("prompt_id")
-    prompt   Prompt? @relation(fields: [promptId], onDelete: SetNull, references: [id])
+    modelParameters  Json?
+    input            Json?
+    output           Json?
+    promptTokens     Int     @default(0) @map("prompt_tokens")
+    completionTokens Int     @default(0) @map("completion_tokens")
+    totalTokens      Int     @default(0) @map("total_tokens")
+    unit             String?
+
+    // User provided cost at ingestion
+    inputCost  Decimal? @map("input_cost")
+    outputCost Decimal? @map("output_cost")
+    totalCost  Decimal? @map("total_cost")
+
+    // Calculated cost
+    calculatedInputCost  Decimal? @map("calculated_input_cost")
+    calculatedOutputCost Decimal? @map("calculated_output_cost")
+    calculatedTotalCost  Decimal? @map("calculated_total_cost")
+
+    completionStartTime DateTime?     @map("completion_start_time")
+    project             Project       @relation(fields: [projectId], references: [id], onDelete: Cascade)
+    derivedDatasetItems DatasetItem[]
+
+    promptId String? @map("prompt_id") // no fk constraint, prompt can be deleted
 
     @@unique([id, projectId])
     @@index([projectId, internalModel, startTime, unit])
@@ -326,11 +338,10 @@ model Observation {
     @@index([type])
     @@index(startTime)
     @@index(createdAt)
-    @@index(updatedAt)
     @@index(projectId)
-    @@index(parentObservationId)
     @@index(model)
     @@index(internalModel)
+    @@index([projectId, promptId])
     @@index(promptId)
     @@index([projectId, startTime, type])
     @@map("observations")
@@ -343,7 +354,7 @@ view ObservationView {
     traceId             String?          @map("trace_id")
     projectId           String           @map("project_id")
     type                ObservationType
-    startTime           DateTime         @default(now()) @map("start_time")
+    startTime           DateTime         @map("start_time")
     endTime             DateTime?        @map("end_time")
     name                String?
     metadata            Json?
@@ -351,7 +362,8 @@ view ObservationView {
     level               ObservationLevel @default(DEFAULT)
     statusMessage       String?          @map("status_message")
     version             String?
-    createdAt           DateTime         @default(now()) @map("created_at")
+    createdAt           DateTime         @map("created_at")
+    updatedAt           DateTime         @map("updated_at")
 
     // GENERATION ONLY
     model               String?
@@ -401,14 +413,14 @@ model Score {
     projectId     String         @map("project_id")
     project       Project        @relation(fields: [projectId], references: [id], onDelete: Cascade)
     name          String
-    value         Float
+    value         Float? // always defined if data type is NUMERIC or BOOLEAN, optional for CATEGORICAL
     source        ScoreSource
     authorUserId  String?        @map("author_user_id")
     comment       String?
     traceId       String         @map("trace_id")
     observationId String?        @map("observation_id")
     configId      String?        @map("config_id")
-    stringValue   String?        @map("string_value")
+    stringValue   String?        @map("string_value") // always defined if data type is CATEGORICAL or BOOLEAN, null for NUMERIC
     createdAt     DateTime       @default(now()) @map("created_at")
     updatedAt     DateTime       @default(now()) @updatedAt @map("updated_at")
     dataType      ScoreDataType  @default(NUMERIC) @map("data_type")
@@ -419,13 +431,13 @@ model Score {
     @@index(timestamp)
     @@index([value])
     @@index([projectId])
+    @@index([projectId, name])
     @@index([authorUserId])
     @@index([configId])
     @@index([traceId], type: Hash)
     @@index([observationId], type: Hash)
     @@index([source])
     @@index([createdAt])
-    @@index([updatedAt])
     @@map("scores")
 }
 
@@ -463,29 +475,7 @@ model ScoreConfig {
 enum ScoreDataType {
     CATEGORICAL
     NUMERIC
-}
-
-enum PricingUnit {
-    PER_1000_TOKENS
-    PER_1000_CHARS
-}
-
-enum TokenType {
-    PROMPT
-    COMPLETION
-    TOTAL
-}
-
-model Pricing {
-    id          String      @id @default(cuid())
-    modelName   String      @map("model_name")
-    pricingUnit PricingUnit @default(PER_1000_TOKENS) @map("pricing_unit")
-    price       Decimal
-    currency    String      @default("USD")
-    tokenType   TokenType   @map("token_type")
-
-    @@index(modelName)
-    @@map("pricings")
+    BOOLEAN
 }
 
 model CronJobs {
@@ -532,6 +522,7 @@ model DatasetItem {
     updatedAt           DateTime          @default(now()) @updatedAt @map("updated_at")
     datasetRunItems     DatasetRunItems[]
 
+    @@index([sourceTraceId], type: Hash)
     @@index([sourceObservationId], type: Hash)
     @@index([datasetId], type: Hash)
     @@index([createdAt])
@@ -563,17 +554,15 @@ model DatasetRuns {
 }
 
 model DatasetRunItems {
-    id            String       @id @default(cuid())
-    datasetRunId  String       @map("dataset_run_id")
-    datasetRun    DatasetRuns  @relation(fields: [datasetRunId], references: [id], onDelete: Cascade)
-    datasetItemId String       @map("dataset_item_id")
-    datasetItem   DatasetItem  @relation(fields: [datasetItemId], references: [id], onDelete: Cascade)
-    traceId       String       @map("trace_id")
-    trace         Trace        @relation(fields: [traceId], references: [id], onDelete: Cascade)
-    observationId String?      @map("observation_id")
-    observation   Observation? @relation(fields: [observationId], references: [id], onDelete: Cascade)
-    createdAt     DateTime     @default(now()) @map("created_at")
-    updatedAt     DateTime     @default(now()) @updatedAt @map("updated_at")
+    id            String      @id @default(cuid())
+    datasetRunId  String      @map("dataset_run_id")
+    datasetRun    DatasetRuns @relation(fields: [datasetRunId], references: [id], onDelete: Cascade)
+    datasetItemId String      @map("dataset_item_id")
+    datasetItem   DatasetItem @relation(fields: [datasetItemId], references: [id], onDelete: Cascade)
+    traceId       String      @map("trace_id")
+    observationId String?     @map("observation_id")
+    createdAt     DateTime    @default(now()) @map("created_at")
+    updatedAt     DateTime    @default(now()) @updatedAt @map("updated_at")
 
     @@index([datasetRunId], type: Hash)
     @@index([datasetItemId], type: Hash)
@@ -609,15 +598,14 @@ model Prompt {
 
     createdBy String @map("created_by")
 
-    prompt      Json
-    name        String
-    version     Int
-    type        String        @default("text")
-    isActive    Boolean?      @map("is_active") // Deprecated. To be removed once 'production' labels work as expected.
-    config      Json          @default("{}")
-    tags        String[]      @default([])
-    labels      String[]      @default([])
-    Observation Observation[]
+    prompt   Json
+    name     String
+    version  Int
+    type     String   @default("text")
+    isActive Boolean? @map("is_active") // Deprecated. To be removed once 'production' labels work as expected.
+    config   Json     @default("{}")
+    tags     String[] @default([])
+    labels   String[] @default([])
 
     @@unique([projectId, name, version])
     @@index([projectId, name, version])
@@ -688,6 +676,7 @@ model EvalTemplate {
     version          Int
     prompt           String
     model            String
+    provider         String
     modelParams      Json               @map("model_params")
     vars             String[]           @default([])
     outputSchema     Json               @map("output_schema")
@@ -801,3 +790,27 @@ model PosthogIntegration {
     @@index([projectId])
     @@map("posthog_integrations")
 }
+
+model BatchExport {
+    id        String   @id @default(cuid())
+    createdAt DateTime @default(now()) @map("created_at")
+    updatedAt DateTime @default(now()) @updatedAt @map("updated_at")
+
+    projectId String  @map("project_id")
+    project   Project @relation(fields: [projectId], references: [id], onDelete: Cascade)
+    userId    String  @map("user_id")
+
+    finishedAt DateTime? @map("finished_at")
+    expiresAt  DateTime? @map("expires_at")
+
+    name   String
+    status String
+    query  Json
+    format String
+    url    String?
+    log    String?
+
+    @@index([projectId, userId])
+    @@index([status])
+    @@map("batch_exports")
+}

packages/shared/prisma/seed.ts

@@ -4,6 +4,7 @@ import {
   type Prisma,
   ObservationType,
   ScoreSource,
+  ScoreDataType,
 } from "../src/index";
 import { hash } from "bcryptjs";
 import { parseArgs } from "node:util";
@@ -11,11 +12,16 @@ import { parseArgs } from "node:util";
 import { chunk } from "lodash";
 import { v4 } from "uuid";
 import { ModelUsageUnit } from "../src";
-import { getDisplaySecretKey, hashSecretKey } from "../src/server/auth";
+import { getDisplaySecretKey, hashSecretKey } from "../src/server";
 import { encrypt } from "../src/encryption";
 
 const LOAD_TRACE_VOLUME = 10_000;
 
+type ConfigCategory = {
+  label: string;
+  value: number;
+};
+
 const options = {
   environment: { type: "string" },
 } as const;
@@ -39,6 +45,7 @@ async function main() {
       name: "Demo User",
       email: "demo@langfuse.com",
       password: await hash("password", 12),
+      image: "https://static.langfuse.com/langfuse-dev%2Fexample-avatar.png",
     },
   });
 
@@ -144,6 +151,11 @@ async function main() {
       });
     }
 
+    const configIdsAndNames = await generateConfigsForProject([
+      project1,
+      project2,
+    ]);
+
     const promptIds = await generatePromptsForProject([project1, project2]);
 
     const envTags = [null, "development", "staging", "production"];
@@ -157,7 +169,8 @@ async function main() {
       colorTags,
       project1,
       project2,
-      promptIds
+      promptIds,
+      configIdsAndNames
     );
 
     console.log(
@@ -176,6 +189,7 @@ async function main() {
           secretKey: encrypt(OPENAI_API_KEY),
           displaySecretKey: getDisplaySecretKey(OPENAI_API_KEY),
           provider: "openai",
+          adapter: "openai",
         },
       });
     } else {
@@ -201,6 +215,7 @@ async function main() {
           "Please evaluate the toxicity of the following text {{input}} {{output}}",
         model: "gpt-3.5-turbo",
         vars: ["input", "output"],
+        provider: "openai",
         outputSchema: {
           score: "provide a score between 0 and 1",
           reasoning: "one sentence reasoning for the score",
@@ -367,11 +382,10 @@ async function uploadObjects(
   });
 
   for (let i = 0; i < promises.length; i++) {
-    process.stdout.clearLine(0);
-    process.stdout.cursorTo(0);
-    process.stdout.write(
-      `Seeding of Sessions ${(i / promises.length) * 100}% complete`
-    );
+    if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
+      console.log(
+        `Seeding of Sessions ${((i + 1) / promises.length) * 100}% complete`
+      );
     await promises[i];
   }
 
@@ -385,11 +399,10 @@ async function uploadObjects(
     );
   });
   for (let i = 0; i < promises.length; i++) {
-    process.stdout.clearLine(0);
-    process.stdout.cursorTo(0);
-    process.stdout.write(
-      `Seeding of Traces ${(i / promises.length) * 100}% complete`
-    );
+    if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
+      console.log(
+        `Seeding of Traces ${((i + 1) / promises.length) * 100}% complete`
+      );
     await promises[i];
   }
 
@@ -403,11 +416,10 @@ async function uploadObjects(
   });
 
   for (let i = 0; i < promises.length; i++) {
-    process.stdout.clearLine(0);
-    process.stdout.cursorTo(0);
-    process.stdout.write(
-      `Seeding of Observations ${(i / promises.length) * 100}% complete`
-    );
+    if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
+      console.log(
+        `Seeding of Observations ${((i + 1) / promises.length) * 100}% complete`
+      );
     await promises[i];
   }
 
@@ -421,11 +433,10 @@ async function uploadObjects(
   });
 
   for (let i = 0; i < promises.length; i++) {
-    process.stdout.clearLine(0);
-    process.stdout.cursorTo(0);
-    process.stdout.write(
-      `Seeding of Events ${(i / promises.length) * 100}% complete`
-    );
+    if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
+      console.log(
+        `Seeding of Events ${((i + 1) / promises.length) * 100}% complete`
+      );
     await promises[i];
   }
 
@@ -438,11 +449,10 @@ async function uploadObjects(
     );
   });
   for (let i = 0; i < promises.length; i++) {
-    process.stdout.clearLine(0);
-    process.stdout.cursorTo(0);
-    process.stdout.write(
-      `Seeding of Scores ${(i / promises.length) * 100}% complete`
-    );
+    if (i + 1 >= promises.length || i % Math.ceil(promises.length / 10) === 0)
+      console.log(
+        `Seeding of Scores ${((i + 1) / promises.length) * 100}% complete`
+      );
     await promises[i];
   }
 }
@@ -453,13 +463,23 @@ function createObjects(
   colorTags: (string | null)[],
   project1: Project,
   project2: Project,
-  promptIds: Map<string, string[]>
+  promptIds: Map<string, string[]>,
+  configParams: Map<
+    string,
+    {
+      name: string;
+      id: string;
+      dataType: ScoreDataType;
+      categories: ConfigCategory[] | null;
+    }[]
+  >
 ) {
   const traces: Prisma.TraceCreateManyInput[] = [];
   const observations: Prisma.ObservationCreateManyInput[] = [];
   const scores: Prisma.ScoreCreateManyInput[] = [];
   const sessions: Prisma.TraceSessionCreateManyInput[] = [];
   const events: Prisma.ObservationCreateManyInput[] = [];
+  const configs: Prisma.ScoreConfigCreateManyInput[] = [];
 
   for (let i = 0; i < traceVolume; i++) {
     // print progress to console with a progress bar that refreshes every 10 iterations
@@ -511,17 +531,49 @@ function createObjects(
 
     traces.push(trace);
 
+    const configArray = configParams.get(projectId) ?? [];
+    const randomIndex = Math.floor(Math.random() * 3);
+    const config =
+      configArray.length >= randomIndex - 1 && configArray[randomIndex];
+    const {
+      name: annotationScoreName,
+      id: configId,
+      dataType,
+      categories,
+    } = config || {
+      name: "manual-score",
+      id: undefined,
+      dataType: ScoreDataType.NUMERIC,
+      categories: null,
+    };
+
+    const value = Math.floor(Math.random() * 2);
+    const scoreNumericAndStringValue = {
+      ...(dataType === ScoreDataType.NUMERIC && { value }),
+      ...(dataType === ScoreDataType.CATEGORICAL && {
+        value,
+        stringValue: categories?.find((category) => category.value === value)
+          ?.label,
+      }),
+      ...(dataType === ScoreDataType.BOOLEAN && {
+        value,
+        stringValue: value === 1 ? "True" : "False",
+      }),
+    };
+
     const traceScores = [
       ...(Math.random() > 0.5
         ? [
             {
               traceId: trace.id,
-              name: "manual-score",
-              value: Math.floor(Math.random() * 3) - 1,
+              name: annotationScoreName,
               timestamp: traceTs,
               source: ScoreSource.ANNOTATION,
               projectId,
               authorUserId: `user-${i}`,
+              dataType,
+              ...scoreNumericAndStringValue,
+              ...(configId ? { configId } : {}),
             },
           ]
         : []),
@@ -534,6 +586,21 @@ function createObjects(
               timestamp: traceTs,
               source: ScoreSource.API,
               projectId,
+              dataType: ScoreDataType.NUMERIC,
+            },
+          ]
+        : []),
+      ...(Math.random() < 0.8
+        ? [
+            {
+              traceId: trace.id,
+              name: "Completeness",
+              timestamp: traceTs,
+              source: ScoreSource.API,
+              projectId,
+              dataType: ScoreDataType.CATEGORICAL,
+              stringValue:
+                Math.floor(Math.random() * 2) === 1 ? "Fully" : "Partially",
             },
           ]
         : []),
@@ -772,6 +839,7 @@ function createObjects(
     traces,
     observations,
     scores,
+    configs,
     sessions: uniqueSessions,
     events,
   };
@@ -954,3 +1022,97 @@ async function generatePrompts(project: Project) {
   }
   return promptIds;
 }
+
+async function generateConfigsForProject(projects: Project[]) {
+  const projectIdsToConfigs: Map<
+    string,
+    {
+      name: string;
+      id: string;
+      dataType: ScoreDataType;
+      categories: ConfigCategory[] | null;
+    }[]
+  > = new Map();
+
+  await Promise.all(
+    projects.map(async (project) => {
+      const configNameAndId = await generateConfigs(project);
+      projectIdsToConfigs.set(project.id, configNameAndId);
+    })
+  );
+  return projectIdsToConfigs;
+}
+
+async function generateConfigs(project: Project) {
+  const configNameAndId: {
+    name: string;
+    id: string;
+    dataType: ScoreDataType;
+    categories: ConfigCategory[] | null;
+  }[] = [];
+
+  const configs = [
+    {
+      id: `config-${v4()}`,
+      name: "manual-score",
+      dataType: ScoreDataType.NUMERIC,
+      projectId: project.id,
+      isArchived: false,
+    },
+    {
+      id: `config-${v4()}`,
+      projectId: project.id,
+      name: "Accuracy",
+      dataType: ScoreDataType.CATEGORICAL,
+      categories: [
+        { label: "Incorrect", value: 0 },
+        { label: "Partially Correct", value: 1 },
+        { label: "Correct", value: 2 },
+      ],
+      isArchived: false,
+    },
+    {
+      id: `config-${v4()}`,
+      projectId: project.id,
+      name: "Toxicity",
+      dataType: ScoreDataType.BOOLEAN,
+      categories: [
+        { label: "True", value: 1 },
+        { label: "False", value: 0 },
+      ],
+      description:
+        "Used to indicate if text was harmful or offensive in nature.",
+      isArchived: false,
+    },
+  ];
+
+  for (const config of configs) {
+    await prisma.scoreConfig.upsert({
+      where: {
+        id_projectId: {
+          projectId: config.projectId,
+          id: config.id,
+        },
+      },
+      create: {
+        id: config.id,
+        projectId: config.projectId,
+        name: config.name,
+        dataType: config.dataType,
+        categories: config.categories,
+        isArchived: config.isArchived,
+      },
+      update: {
+        id: config.id,
+      },
+    });
+    configNameAndId.push({
+      name: config.name,
+      id: config.id,
+      dataType: config.dataType,
+      categories: config.categories ?? null,
+    });
+  }
+
+  return configNameAndId;
+}

packages/shared/scripts/cleanup.sql

@@ -0,0 +1,7 @@
+DO $$
+BEGIN
+ IF EXISTS (SELECT 1 FROM information_schema.tables WHERE table_name = '_prisma_migrations') THEN
+  DELETE FROM _prisma_migrations
+  WHERE migration_name IN ('20240606090858_pricings_add_latest_gemini_models', '20240530212419_model_price_anthropic_via_google_vertex', '20240604133340_backfill_manual_scores');
+ END IF;
+END $$;

packages/shared/src/errors/InternalServerError.ts

@@ -0,0 +1,7 @@
+import { BaseError } from "./BaseError";
+
+export class InternalServerError extends BaseError {
+  constructor(description = "Internal Server Error") {
+    super("InternalServerError", 500, description, true);
+  }
+}

packages/shared/src/errors/InvalidRequestError.ts

@@ -0,0 +1,7 @@
+import { BaseError } from "./BaseError";
+
+export class InvalidRequestError extends BaseError {
+  constructor(description = "Invalid Request Error") {
+    super("InvalidRequestError", 400, description, true);
+  }
+}

packages/shared/src/errors/ValidationError.ts

@@ -1,7 +0,0 @@
-import { BaseError } from "./BaseError";
-
-export class ValidationError extends BaseError {
-  constructor(description = "Validation Error") {
-    super("ValidationError", 400, description, true);
-  }
-}

packages/shared/src/errors/index.ts

@@ -1,7 +1,8 @@
 export { BaseError } from "./BaseError";
 export { LangfuseNotFoundError } from "./NotFoundError";
-export { ValidationError } from "./ValidationError";
+export { InvalidRequestError } from "./InvalidRequestError";
 export { UnauthorizedError } from "./UnauthorizedError";
 export { ForbiddenError } from "./ForbiddenError";
 export { MethodNotAllowedError } from "./MethodNotAllowedError";
 export { ApiError } from "./ApiError";
+export { InternalServerError } from "./InternalServerError";

packages/shared/src/features/annotation/types.ts

@@ -0,0 +1,59 @@
+import z from "zod";
+import { type ScoreDataType } from "../../db";
+
+const NUMERIC: ScoreDataType = "NUMERIC";
+const CATEGORICAL: ScoreDataType = "CATEGORICAL";
+const BOOLEAN: ScoreDataType = "BOOLEAN";
+
+export const availableDataTypes = [NUMERIC, CATEGORICAL, BOOLEAN] as const;
+
+const NumericData = z.object({
+  value: z.number(),
+  stringValue: z.undefined().nullish(),
+  dataType: z.literal("NUMERIC"),
+});
+
+const CategoricalData = z.object({
+  value: z.number().optional().nullish(),
+  stringValue: z.string(),
+  dataType: z.literal("CATEGORICAL"),
+});
+
+const BooleanData = z.object({
+  value: z.number(),
+  stringValue: z.string(),
+  dataType: z.literal("BOOLEAN"),
+});
+
+const CreateAnnotationScoreBase = z.object({
+  name: z.string(),
+  projectId: z.string(),
+  traceId: z.string(),
+  configId: z.string().optional(),
+  observationId: z.string().optional(),
+  comment: z.string().optional().nullish(),
+});
+
+const UpdateAnnotationScoreBase = CreateAnnotationScoreBase.extend({
+  id: z.string(),
+});
+
+/**
+ * CreateAnnotationScoreData is only used for annotation scores created via the UI.
+ * For langfuse score types please refer to `web/src/features/public-api/types/scores.ts`
+ */
+export const CreateAnnotationScoreData = z.discriminatedUnion("dataType", [
+  CreateAnnotationScoreBase.merge(NumericData),
+  CreateAnnotationScoreBase.merge(CategoricalData),
+  CreateAnnotationScoreBase.merge(BooleanData),
+]);
+
+/**
+ * UpdateAnnotationScoreData is only used for annotation scores updated via the UI
+ * For langfuse score types please refer to `web/src/features/public-api/types/scores.ts`
+ */
+export const UpdateAnnotationScoreData = z.discriminatedUnion("dataType", [
+  UpdateAnnotationScoreBase.merge(NumericData),
+  UpdateAnnotationScoreBase.merge(CategoricalData),
+  UpdateAnnotationScoreBase.merge(BooleanData),
+]);

packages/shared/src/features/batchExport/types.ts

@@ -0,0 +1,73 @@
+import z from "zod";
+
+import { BatchExport } from "@prisma/client";
+
+import { singleFilter } from "../../interfaces/filters";
+import { orderBy } from "../../interfaces/orderBy";
+
+export enum BatchExportStatus {
+  QUEUED = "QUEUED",
+  PROCESSING = "PROCESSING",
+  COMPLETED = "COMPLETED",
+  FAILED = "FAILED",
+}
+
+export enum BatchExportFileFormat {
+  JSON = "JSON",
+  CSV = "CSV",
+}
+
+export enum BatchExportTableName {
+  Sessions = "sessions",
+}
+
+export const exportOptions: Record<
+  BatchExportFileFormat,
+  {
+    label: string;
+    extension: string;
+    fileType: string;
+  }
+> = {
+  CSV: { label: "CSV", extension: "csv", fileType: "text/csv" },
+  JSON: { label: "JSON", extension: "json", fileType: "application/json" },
+} as const;
+
+export const BatchExportQuerySchema = z.object({
+  tableName: z.nativeEnum(BatchExportTableName),
+  filter: z.array(singleFilter).nullable(),
+  orderBy,
+  limit: z.number().optional(),
+  page: z.number().optional(),
+});
+
+export type BatchExportQueryType = z.infer<typeof BatchExportQuerySchema>;
+
+export const CreateBatchExportSchema = z.object({
+  projectId: z.string(),
+  name: z.string(),
+  query: BatchExportQuerySchema,
+  format: z.nativeEnum(BatchExportFileFormat),
+});
+
+export const BatchExportSchema = z.object({
+  id: z.string(),
+  createdAt: z.date(),
+  updatedAt: z.date(),
+  projectId: z.string(),
+  userId: z.string(),
+  finishedAt: z.date().nullable(),
+  expiresAt: z.date().nullable(),
+  name: z.string(),
+  status: z.nativeEnum(BatchExportStatus),
+  query: BatchExportQuerySchema,
+  format: z.nativeEnum(BatchExportFileFormat),
+  url: z.string().nullable(),
+  log: z.string().nullable(),
+});
+
+// Ensure that zod type matches the Prisma type
+export type BatchExportType =
+  z.infer<typeof BatchExportSchema> extends BatchExport
+    ? z.infer<typeof BatchExportSchema>
+    : never;

packages/shared/src/features/evals/types.ts

@@ -1,5 +1,5 @@
 import z from "zod";
-import { ModelProvider, UIModelParams } from "../..";
+import { LLMAdapter } from "../..";
 
 export const langfuseObjects = [
   "trace",
@@ -77,38 +77,15 @@ export const availableEvalVariables = [
   },
 ];
 
-export const evalLLMModels: UIModelParams[] = [
-  {
-    provider: { value: ModelProvider.OpenAI, enabled: true },
-    model: { value: "gpt-3.5-turbo", enabled: true },
-    temperature: { value: 1, enabled: true },
-    maxTemperature: { value: 2, enabled: true },
-    max_tokens: { value: 256, enabled: true },
-    top_p: { value: 1, enabled: true },
-  },
-  {
-    provider: { value: ModelProvider.OpenAI, enabled: true },
-    model: { value: "gpt-4-turbo-preview", enabled: true },
-    temperature: { value: 1, enabled: true },
-    maxTemperature: { value: 2, enabled: true },
-    max_tokens: { value: 256, enabled: true },
-    top_p: { value: 1, enabled: true },
-  },
-  {
-    provider: { value: ModelProvider.OpenAI, enabled: true },
-    model: { value: "gpt-4o", enabled: true },
-    temperature: { value: 1, enabled: true },
-    maxTemperature: { value: 2, enabled: true },
-    max_tokens: { value: 256, enabled: true },
-    top_p: { value: 1, enabled: true },
-  },
-] as const;
-
-export const EvalModelNames = z.enum([
-  "gpt-3.5-turbo",
-  "gpt-4-turbo-preview",
-  "gpt-4o",
-]);
+export const evalModelsByAdapter: Record<LLMAdapter, string[]> = {
+  [LLMAdapter.OpenAI]: [
+    "gpt-4o",
+    "gpt-4-turbo-preview",
+    "gpt-3.5-turbo",
+  ] as const,
+  [LLMAdapter.Anthropic]: [] as const,
+  [LLMAdapter.Azure]: [] as const,
+};
 
 export const OutputSchema = z.object({
   reasoning: z.string(),

packages/shared/src/features/ingestion/types.ts

@@ -1,9 +1,9 @@
 import lodash from "lodash";
 import { z } from "zod";
 
-import { ModelUsageUnit } from "@langfuse/shared";
-import { NonEmptyString, jsonSchema } from "@/src/utils/zod";
-import { ObservationLevel } from "@langfuse/shared";
+import { NonEmptyString, jsonSchema } from "../../utils/zod";
+import { ModelUsageUnit } from "../../constants";
+import { ObservationLevel } from "@prisma/client";
 
 export const Usage = z.object({
   input: z.number().int().nullish(),
@@ -28,7 +28,7 @@ const MixedUsage = z.object({
   totalCost: z.number().nullish(),
 });
 
-export const stringDate = z.string().datetime({ offset: true }).nullish();
+export const stringDateTime = z.string().datetime({ offset: true }).nullish();
 
 export const usage = MixedUsage.nullish()
   // transform mixed usage model input to new one
@@ -58,7 +58,7 @@ export const usage = MixedUsage.nullish()
 
 export const TraceBody = z.object({
   id: z.string().nullish(),
-  timestamp: stringDate,
+  timestamp: stringDateTime,
   name: z.string().nullish(),
   externalId: z.string().nullish(),
   input: jsonSchema.nullish(),
@@ -75,7 +75,7 @@ export const TraceBody = z.object({
 export const OptionalObservationBody = z.object({
   traceId: z.string().nullish(),
   name: z.string().nullish(),
-  startTime: stringDate,
+  startTime: stringDateTime,
   metadata: jsonSchema.nullish(),
   input: jsonSchema.nullish(),
   output: jsonSchema.nullish(),
@@ -94,22 +94,28 @@ export const UpdateEventEvent = OptionalObservationBody.extend({
 });
 
 export const CreateSpanBody = CreateEventEvent.extend({
-  endTime: stringDate,
+  endTime: stringDateTime,
 });
 
 export const UpdateSpanBody = UpdateEventEvent.extend({
-  endTime: stringDate,
+  endTime: stringDateTime,
 });
 
 export const CreateGenerationBody = CreateSpanBody.extend({
-  completionStartTime: stringDate,
+  completionStartTime: stringDateTime,
   model: z.string().nullish(),
   modelParameters: z
     .record(
       z.string(),
       z
-        .union([z.string(), z.number(), z.boolean(), z.array(z.string())])
-        .nullish(),
+        .union([
+          z.string(),
+          z.number(),
+          z.boolean(),
+          z.array(z.string()),
+          z.record(z.string()),
+        ])
+        .nullish()
     )
     .nullish(),
   usage: usage,
@@ -124,14 +130,20 @@ export const CreateGenerationBody = CreateSpanBody.extend({
 });
 
 export const UpdateGenerationBody = UpdateSpanBody.extend({
-  completionStartTime: stringDate,
+  completionStartTime: stringDateTime,
   model: z.string().nullish(),
   modelParameters: z
     .record(
       z.string(),
       z
-        .union([z.string(), z.number(), z.boolean(), z.array(z.string())])
-        .nullish(),
+        .union([
+          z.string(),
+          z.number(),
+          z.boolean(),
+          z.array(z.string()),
+          z.record(z.string()),
+        ])
+        .nullish()
     )
     .nullish(),
   usage: usage,
@@ -145,22 +157,82 @@ export const UpdateGenerationBody = UpdateSpanBody.extend({
   return false;
 });
 
-export const ScoreBody = z.object({
+const BaseScoreBody = z.object({
   id: z.string().nullish(),
   name: NonEmptyString,
-  value: z.number(),
   traceId: z.string(),
   observationId: z.string().nullish(),
   comment: z.string().nullish(),
 });
 
+/**
+ * ScoreBody exactly mirrors `PostScoresBody` in the public API. Please refer there for source of truth.
+ */
+export const ScoreBody = z
+  .discriminatedUnion("dataType", [
+    BaseScoreBody.merge(
+      z.object({
+        value: z.number(),
+        dataType: z.literal("NUMERIC"),
+        configId: z.string().nullish(),
+      })
+    ),
+    BaseScoreBody.merge(
+      z.object({
+        value: z.string(),
+        dataType: z.literal("CATEGORICAL"),
+        configId: z.string().nullish(),
+      })
+    ),
+    BaseScoreBody.merge(
+      z.object({
+        value: z.number().refine((val) => val === 0 || val === 1, {
+          message: "Value must be either 0 or 1",
+        }),
+        dataType: z.literal("BOOLEAN"),
+        configId: z.string().nullish(),
+      })
+    ),
+    BaseScoreBody.merge(
+      z.object({
+        value: z.union([z.string(), z.number()]),
+        dataType: z.undefined(),
+        configId: z.string().nullish(),
+      })
+    ),
+  ])
+  .superRefine((data, ctx) => {
+    if (data.dataType) {
+      if (typeof data.value === "number") {
+        if (data.dataType === "CATEGORICAL") {
+          ctx.addIssue({
+            code: z.ZodIssueCode.custom,
+            message: `Value must be a string for data type ${data.dataType}`,
+          });
+        }
+      } else if (typeof data.value === "string") {
+        if (data.dataType === "NUMERIC") {
+          ctx.addIssue({
+            code: z.ZodIssueCode.custom,
+            message: `Value must be a number for data type ${data.dataType}`,
+          });
+        } else if (data.dataType === "BOOLEAN") {
+          ctx.addIssue({
+            code: z.ZodIssueCode.custom,
+            message: `Value must number equal to either 0 or 1 for data type ${data.dataType}`,
+          });
+        }
+      }
+    }
+  });
+
 // LEGACY, only required for backwards compatibility
 export const LegacySpanPostSchema = z.object({
   id: z.string().nullish(),
   traceId: z.string().nullish(),
   name: z.string().nullish(),
-  startTime: stringDate,
-  endTime: stringDate,
+  startTime: stringDateTime,
+  endTime: stringDateTime,
   metadata: jsonSchema.nullish(),
   input: jsonSchema.nullish(),
   output: jsonSchema.nullish(),
@@ -174,8 +246,8 @@ export const LegacySpanPatchSchema = z.object({
   spanId: z.string(),
   traceId: z.string().nullish(),
   name: z.string().nullish(),
-  startTime: stringDate,
-  endTime: stringDate,
+  startTime: stringDateTime,
+  endTime: stringDateTime,
   metadata: jsonSchema.nullish(),
   input: jsonSchema.nullish(),
   output: jsonSchema.nullish(),
@@ -188,14 +260,14 @@ export const LegacyGenerationsCreateSchema = z.object({
   id: z.string().nullish(),
   traceId: z.string().nullish(),
   name: z.string().nullish(),
-  startTime: stringDate,
-  endTime: stringDate,
-  completionStartTime: stringDate,
+  startTime: stringDateTime,
+  endTime: stringDateTime,
+  completionStartTime: stringDateTime,
   model: z.string().nullish(),
   modelParameters: z
     .record(
       z.string(),
-      z.union([z.string(), z.number(), z.boolean()]).nullish(),
+      z.union([z.string(), z.number(), z.boolean()]).nullish()
     )
     .nullish(),
   prompt: jsonSchema.nullish(),
@@ -212,14 +284,14 @@ export const LegacyGenerationPatchSchema = z.object({
   generationId: z.string(),
   traceId: z.string().nullish(),
   name: z.string().nullish(),
-  startTime: stringDate,
-  endTime: stringDate,
-  completionStartTime: stringDate,
+  startTime: stringDateTime,
+  endTime: stringDateTime,
+  completionStartTime: stringDateTime,
   model: z.string().nullish(),
   modelParameters: z
     .record(
       z.string(),
-      z.union([z.string(), z.number(), z.boolean()]).nullish(),
+      z.union([z.string(), z.number(), z.boolean()]).nullish()
     )
     .nullish(),
   prompt: jsonSchema.nullish(),
@@ -236,14 +308,14 @@ export const LegacyObservationBody = z.object({
   traceId: z.string().nullish(),
   type: z.enum(["GENERATION", "SPAN", "EVENT"]),
   name: z.string().nullish(),
-  startTime: stringDate,
-  endTime: stringDate,
-  completionStartTime: stringDate,
+  startTime: stringDateTime,
+  endTime: stringDateTime,
+  completionStartTime: stringDateTime,
   model: z.string().nullish(),
   modelParameters: z
     .record(
       z.string(),
-      z.union([z.string(), z.number(), z.boolean()]).nullish(),
+      z.union([z.string(), z.number(), z.boolean()]).nullish()
     )
     .nullish(),
   input: jsonSchema.nullish(),

packages/shared/src/filterToPrisma.ts

@@ -1,8 +1,5 @@
 import { Prisma } from "@prisma/client";
-import {
-  ColumnDefinition,
-  type TableNames as TableName,
-} from "./interfaces/tableDefinition";
+import { ColumnDefinition, type TableNames } from "./tableDefinitions";
 import { FilterState } from "./types";
 import { filterOperators } from "./interfaces/filters";
 
@@ -27,7 +24,7 @@ const arrayOperatorReplacements = {
 export function tableColumnsToSqlFilterAndPrefix(
   filters: FilterState,
   tableColumns: ColumnDefinition[],
-  table: TableName
+  table: TableNames
 ): Prisma.Sql {
   const sql = tableColumnsToSqlFilter(filters, tableColumns, table);
   if (sql === Prisma.empty) {
@@ -43,7 +40,7 @@ export function tableColumnsToSqlFilterAndPrefix(
 export function tableColumnsToSqlFilter(
   filters: FilterState,
   tableColumns: ColumnDefinition[],
-  table: TableName
+  table: TableNames
 ): Prisma.Sql {
   const internalFilters = filters.map((filter) => {
     // Get column definition to map column to internal name, e.g. "t.id"
@@ -153,7 +150,7 @@ export function tableColumnsToSqlFilter(
 
 const castValueToPostgresTypes = (
   column: ColumnDefinition,
-  table: TableName
+  table: TableNames
 ) => {
   return column.name === "type" &&
     (table === "observations" ||

packages/shared/src/index.ts

@@ -1,14 +1,18 @@
 export * from "./constants";
+export * from "./queries";
 export * from "./queues";
-export * from "./interfaces/exportTypes";
 export * from "./interfaces/filters";
 export * from "./interfaces/orderBy";
-export * from "./interfaces/tableDefinition";
+export * from "./tableDefinitions";
 export * from "./types";
 export * from "./filterToPrisma";
+export * from "./orderByToPrisma";
 export * from "./tracesTable";
 export * from "./server/auth";
 export * from "./observationsTable";
+export * from "./features/ingestion/types";
+export * from "./utils/zod";
+export * from "./utils/json";
 export { env } from "./env";
 
 // llm api
@@ -17,6 +21,10 @@ export * from "./server/llm/fetchLLMCompletion";
 
 // evals
 export * from "./features/evals/types";
+export * from "./features/batchExport/types";
+
+// annotation
+export * from "./features/annotation/types";
 
 // export db types only
 export * from "@prisma/client";

packages/shared/src/interfaces/exportTypes.ts

@@ -1,19 +0,0 @@
-export const exportFileFormats = ["CSV", "JSON", "OPENAI-JSONL"] as const;
-export type ExportFileFormats = (typeof exportFileFormats)[number];
-
-export const exportOptions: Record<
-  ExportFileFormats,
-  {
-    label: string;
-    extension: string;
-    fileType: string;
-  }
-> = {
-  CSV: { label: "CSV", extension: "csv", fileType: "text/csv" },
-  JSON: { label: "JSON", extension: "json", fileType: "application/json" },
-  "OPENAI-JSONL": {
-    label: "OpenAI JSONL (fine-tuning)",
-    extension: "jsonl",
-    fileType: "application/json",
-  },
-} as const;

packages/shared/src/interfaces/orderBy.ts

@@ -6,3 +6,5 @@ export const orderBy = z
     order: z.enum(["ASC", "DESC"]),
   })
   .nullable();
+
+export type OrderByState = z.infer<typeof orderBy>;

packages/shared/src/observationsTable.ts

@@ -2,7 +2,7 @@ import { ObservationLevel } from "@prisma/client";
 import {
   type OptionsDefinition,
   type ColumnDefinition,
-} from "./interfaces/tableDefinition";
+} from "./tableDefinitions/types";
 
 // to be used server side
 export const observationsTableCols: ColumnDefinition[] = [

packages/shared/src/orderByToPrisma.ts

@@ -1,8 +1,10 @@
-import { type OrderByState } from "@/src/features/orderBy/types";
-import { type ColumnDefinition } from "@langfuse/shared";
-import { Prisma } from "@langfuse/shared";
 import { z } from "zod";
 
+import { Prisma } from "@prisma/client";
+
+import type { ColumnDefinition } from "./tableDefinitions/types";
+import type { OrderByState } from "./interfaces/orderBy";
+
 /**
  * Convert orderBy to SQL ORDER BY clause
  * @param orderBy orderBy state
@@ -11,7 +13,7 @@ import { z } from "zod";
  */
 export function orderByToPrismaSql(
   orderBy: OrderByState,
-  tableColumns: ColumnDefinition[],
+  tableColumns: ColumnDefinition[]
 ): Prisma.Sql {
   if (!orderBy) {
     return Prisma.sql`ORDER BY t.timestamp DESC NULLS LAST`;
@@ -20,7 +22,7 @@ export function orderByToPrismaSql(
   const col = tableColumns.find(
     // TODO: Only use id instead of name.
     // It's less error-prone & decouples data fetching from the human-readable UI labels
-    (c) => c.name === orderBy.column || c.id === orderBy.column,
+    (c) => c.name === orderBy.column || c.id === orderBy.column
   );
 
   if (!col) {
@@ -38,6 +40,6 @@ export function orderByToPrismaSql(
 
   // Both column and order are safe, can use raw SQL
   return Prisma.raw(
-    `ORDER BY ${col.internal} ${order.data} ${orderBy.order === "DESC" ? "NULLS LAST" : "NULLS FIRST"}`,
+    `ORDER BY ${col.internal} ${order.data} ${orderBy.order === "DESC" ? "NULLS LAST" : "NULLS FIRST"}`
   );
 }

packages/shared/src/queries/getSessionTableSQL.ts

@@ -0,0 +1,88 @@
+import { z } from "zod";
+
+import { Prisma } from "@prisma/client";
+import { tableColumnsToSqlFilterAndPrefix } from "../filterToPrisma";
+import { singleFilter } from "../interfaces/filters";
+import { orderBy } from "../interfaces/orderBy";
+import { orderByToPrismaSql } from "../orderByToPrisma";
+import { sessionsViewCols } from "../tableDefinitions/index";
+
+const GetSessionTableSQLParamsSchema = z.object({
+  projectId: z.string(),
+  filter: z.array(singleFilter).nullable(),
+  orderBy: orderBy,
+  page: z.number(),
+  limit: z.number(),
+});
+type GetSessionTableSQLParams = z.infer<typeof GetSessionTableSQLParamsSchema>;
+
+export const getSessionTableSQL = (
+  params: GetSessionTableSQLParams
+): Prisma.Sql => {
+  const { projectId, filter, orderBy, page, limit } =
+    GetSessionTableSQLParamsSchema.parse(params);
+
+  const filterCondition = tableColumnsToSqlFilterAndPrefix(
+    filter ?? [],
+    sessionsViewCols,
+    "sessions"
+  );
+  const orderByCondition = orderByToPrismaSql(orderBy, sessionsViewCols);
+
+  const sql = Prisma.sql`
+      SELECT
+        s.id,
+        s. "created_at" AS "createdAt",
+        s.bookmarked,
+        s.public,
+        t. "userIds",
+        t. "countTraces",
+        o. "sessionDuration",
+        o. "totalCost" AS "totalCost",
+        o. "inputCost" AS "inputCost",
+        o. "outputCost" AS "outputCost",
+        o. "promptTokens" AS "promptTokens",
+        o. "completionTokens" AS "completionTokens",
+        o. "totalTokens" AS "totalTokens",
+        (count(*) OVER ())::int AS "totalCount"
+      FROM
+        trace_sessions AS s
+        LEFT JOIN LATERAL (
+          SELECT
+            t.session_id,
+            MAX(t. "timestamp") AS "max_timestamp",
+            MIN(t. "timestamp") AS "min_timestamp",
+            array_agg(t.id) AS "traceIds",
+            array_agg(DISTINCT t.user_id) AS "userIds",
+            count(t.id)::int AS "countTraces"
+          FROM
+            traces t
+          WHERE
+            t.project_id = ${projectId}
+            AND t.session_id = s.id
+          GROUP BY
+            t.session_id) AS t ON TRUE
+        LEFT JOIN LATERAL (
+          SELECT
+            EXTRACT(EPOCH FROM COALESCE(MAX(o. "end_time"), MAX(o. "start_time"), t. "max_timestamp")) - EXTRACT(EPOCH FROM COALESCE(MIN(o. "start_time"), t. "min_timestamp"))::double precision AS "sessionDuration",
+            SUM(COALESCE(o. "calculated_input_cost", 0)) AS "inputCost",
+            SUM(COALESCE(o. "calculated_output_cost", 0)) AS "outputCost",
+            SUM(COALESCE(o. "calculated_total_cost", 0)) AS "totalCost",
+            SUM(o.prompt_tokens) AS "promptTokens",
+            SUM(o.completion_tokens) AS "completionTokens",
+            SUM(o.total_tokens) AS "totalTokens"
+          FROM
+            observations_view o
+          WHERE
+            o.project_id = ${projectId}
+            AND o.trace_id = ANY (t. "traceIds")) AS o ON TRUE
+      WHERE
+        s. "project_id" = ${projectId}
+        ${filterCondition}
+      ${orderByCondition}
+      LIMIT ${limit}
+      OFFSET ${page * limit}
+    `;
+
+  return sql;
+};

packages/shared/src/queries/index.ts

@@ -0,0 +1 @@
+export { getSessionTableSQL } from "./getSessionTableSQL";

packages/shared/src/queues.ts

@@ -1,36 +1,75 @@
 import { z } from "zod";
 
-export const TraceUpsertEvent = z.object({
+export enum EventName {
+  TraceUpsert = "TraceUpsert",
+  BatchExport = "BatchExport",
+  EvaluationExecution = "EvaluationExecution",
+}
+
+export const BatchExportJobSchema = z.object({
+  projectId: z.string(),
+  batchExportId: z.string(),
+});
+export const TraceUpsertEventSchema = z.object({
   projectId: z.string(),
   traceId: z.string(),
 });
-
 export const EvalExecutionEvent = z.object({
   projectId: z.string(),
   jobExecutionId: z.string(),
 });
 
+export type BatchExportJobType = z.infer<typeof BatchExportJobSchema>;
+export type TraceUpsertEventType = z.infer<typeof TraceUpsertEventSchema>;
+export type EvalExecutionEventType = z.infer<typeof EvalExecutionEvent>;
+
+export const EventBodySchema = z.union([
+  z.object({
+    name: z.literal(EventName.TraceUpsert),
+    payload: z.array(TraceUpsertEventSchema),
+  }),
+  z.object({
+    name: z.literal(EventName.EvaluationExecution),
+    payload: EvalExecutionEvent,
+  }),
+  z.object({
+    name: z.literal(EventName.BatchExport),
+    payload: BatchExportJobSchema,
+  }),
+]);
+export type EventBodyType = z.infer<typeof EventBodySchema>;
+
 export enum QueueName {
   TraceUpsert = "trace-upsert", // Ingestion pipeline adds events on each Trace upsert
   EvaluationExecution = "evaluation-execution-queue", // Worker executes Evals
+  BatchExport = "batch-export-queue",
+  RepeatQueue = "repeat-queue",
 }
 
 export enum QueueJobs {
   TraceUpsert = "trace-upsert",
   EvaluationExecution = "evaluation-execution-job",
+  BatchExportJob = "batch-export-job",
+  EnqueueBatchExportJobs = "enqueue-batch-export-jobs",
 }
 
 export type TQueueJobTypes = {
   [QueueName.TraceUpsert]: {
     timestamp: Date;
     id: string;
-    payload: z.infer<typeof TraceUpsertEvent>;
+    payload: TraceUpsertEventType;
     name: QueueJobs.TraceUpsert;
   };
   [QueueName.EvaluationExecution]: {
     timestamp: Date;
     id: string;
-    payload: z.infer<typeof EvalExecutionEvent>;
+    payload: EvalExecutionEventType;
     name: QueueJobs.EvaluationExecution;
   };
+  [QueueName.BatchExport]: {
+    timestamp: Date;
+    id: string;
+    payload: BatchExportJobType;
+    name: QueueJobs.BatchExportJob;
+  };
 };

packages/shared/src/server/auth.ts

@@ -1,6 +1,7 @@
 import { compare, hash } from "bcryptjs";
 import { randomUUID } from "crypto";
 import * as crypto from "crypto";
+import type { OAuthConfig, OAuthUserConfig } from "next-auth/providers/oauth";
 
 export function generateSecretKey() {
   return `sk-lf-${randomUUID()}`;
@@ -48,3 +49,33 @@ export function createShaHash(privateKey: string, salt: string): string {
 
   return hash;
 }
+
+export interface CustomSSOUser extends Record<string, any> {
+  email: string;
+  id: string;
+  name: string;
+  verified: boolean;
+}
+
+export function CustomSSOProvider<P extends CustomSSOUser>(
+  options: OAuthUserConfig<P>
+): OAuthConfig<P> {
+  return {
+    id: "custom",
+    name: "CustomSSOProvider",
+    type: "oauth",
+    wellKnown: `${options.issuer}/.well-known/openid-configuration`,
+    authorization: { params: { scope: "openid email profile" } }, // overridden by options.authorization to be able to set custom scopes, deep merged with this default
+    checks: ["pkce", "state"],
+    idToken: true,
+    profile(profile) {
+      return {
+        id: profile.sub,
+        name: profile.name,
+        email: profile.email,
+        image: null,
+      };
+    },
+    options,
+  };
+}

packages/shared/src/server/index.ts

@@ -0,0 +1,9 @@
+export * from "./services/S3StorageService";
+export * from "./services/email/projectInvitation/sendProjectInvitationEmail";
+export * from "./services/email/batchExportSuccess/sendBatchExportSuccessEmail";
+export * from "./services/email/passwordReset/sendResetPasswordVerificationRequest";
+export * from "./auth";
+export * from "./llm/fetchLLMCompletion";
+export * from "./llm/types";
+export * from "./utils/DatabaseReadStream";
+export * from "./utils/transforms";

packages/shared/src/server/llm/fetchLLMCompletion.ts

@@ -17,7 +17,7 @@ import {
   ChatMessageRole,
   LLMFunctionCall,
   ModelParams,
-  ModelProvider,
+  LLMAdapter,
 } from "./types";
 import zodToJsonSchema from "zod-to-json-schema";
 import { JsonOutputFunctionsParser } from "langchain/output_parsers";
@@ -27,7 +27,9 @@ type LLMCompletionParams = {
   modelParams: ModelParams;
   functionCall?: LLMFunctionCall;
   callbacks?: BaseCallbackHandler[];
+  baseURL?: string;
   apiKey?: string;
+  maxRetries?: number;
 };
 
 type FetchLLMCompletionParams = LLMCompletionParams & {
@@ -57,7 +59,16 @@ export async function fetchLLMCompletion(
   params: FetchLLMCompletionParams
 ): Promise<string | IterableReadableStream<Uint8Array> | unknown> {
   // the apiKey must never be printed to the console
-  const { messages, modelParams, streaming, callbacks, apiKey } = params;
+  const {
+    messages,
+    modelParams,
+    streaming,
+    callbacks,
+    apiKey,
+    baseURL,
+    maxRetries,
+  } = params;
+
   const finalMessages = messages.map((message) => {
     if (message.role === ChatMessageRole.User)
       return new HumanMessage(message.content);
@@ -68,16 +79,18 @@ export async function fetchLLMCompletion(
   });
 
   let chatModel: ChatOpenAI | ChatAnthropic;
-  if (modelParams.provider === ModelProvider.Anthropic) {
+  if (modelParams.adapter === LLMAdapter.Anthropic) {
     chatModel = new ChatAnthropic({
       anthropicApiKey: apiKey,
+      anthropicApiUrl: baseURL,
       modelName: modelParams.model,
       temperature: modelParams.temperature,
       maxTokens: modelParams.max_tokens,
       topP: modelParams.top_p,
       callbacks,
+      clientOptions: { maxRetries },
     });
-  } else if (modelParams.provider === ModelProvider.OpenAI) {
+  } else if (modelParams.adapter === LLMAdapter.OpenAI) {
     chatModel = new ChatOpenAI({
       openAIApiKey: apiKey,
       modelName: modelParams.model,
@@ -85,15 +98,29 @@ export async function fetchLLMCompletion(
       maxTokens: modelParams.max_tokens,
       topP: modelParams.top_p,
       callbacks,
+      maxRetries,
+      configuration: {
+        baseURL,
+      },
+    });
+  } else if (modelParams.adapter === LLMAdapter.Azure) {
+    chatModel = new ChatOpenAI({
+      azureOpenAIApiKey: apiKey,
+      azureOpenAIBasePath: baseURL,
+      azureOpenAIApiDeploymentName: modelParams.model,
+      azureOpenAIApiVersion: "2024-02-01",
+      temperature: modelParams.temperature,
+      maxTokens: modelParams.max_tokens,
+      topP: modelParams.top_p,
+      callbacks,
+      maxRetries,
     });
   } else {
     // eslint-disable-next-line no-unused-vars
-    const _exhaustiveCheck: never = modelParams;
+    const _exhaustiveCheck: never = modelParams.adapter;
     throw new Error("This model provider is not supported.");
   }
 
-  console.log("Making LLM call with params: ", modelParams);
-
   if (params.functionCall) {
     const functionCallingModel = chatModel.bind({
       functions: [

packages/shared/src/server/llm/types.ts

@@ -1,3 +1,4 @@
+import { LlmApiKeys } from "@prisma/client";
 import z from "zod";
 
 export type PromptVariable = { name: string; value: string; isUsed: boolean };
@@ -9,9 +10,10 @@ export type ChatMessage = {
 
 export type ChatMessageWithId = ChatMessage & { id: string };
 
-export enum ModelProvider {
+export enum LLMAdapter {
   Anthropic = "anthropic",
   OpenAI = "openai",
+  Azure = "azure",
 }
 
 export enum ChatMessageRole {
@@ -19,32 +21,30 @@ export enum ChatMessageRole {
   User = "user",
   Assistant = "assistant",
 }
-export type ModelParams = AnthropicModelParams | OpenAIModelParams;
+
+export type ModelParams = {
+  provider: string;
+  adapter: LLMAdapter;
+  model: string;
+} & ModelConfig;
 
 type RecordWithEnabledFlag<T> = {
   [K in keyof T]: { value: T[K]; enabled: boolean };
 };
 export type UIModelParams = RecordWithEnabledFlag<
-  Required<AnthropicModelParams | OpenAIModelParams> & {
+  Required<ModelParams> & {
     maxTemperature: number;
   }
 >;
 
 // Generic config
 export type ModelConfig = z.infer<typeof ZodModelConfig>;
-
 export const ZodModelConfig = z.object({
   max_tokens: z.coerce.number().optional(),
   temperature: z.coerce.number().optional(),
   top_p: z.coerce.number().optional(),
 });
 
-// OpenAI
-export type OpenAIModelParams = {
-  provider: ModelProvider.OpenAI;
-  model: OpenAIModel;
-} & ModelConfig;
-
 export const openAIModels = [
   "gpt-4o",
   "gpt-4o-2024-05-13",
@@ -64,13 +64,8 @@ export const openAIModels = [
 
 export type OpenAIModel = (typeof openAIModels)[number];
 
-// Anthropic
-export type AnthropicModelParams = {
-  provider: ModelProvider.Anthropic;
-  model: AnthropicModel;
-} & ModelConfig;
-
 export const anthropicModels = [
+  "claude-3-5-sonnet-20240620",
   "claude-3-opus-20240229",
   "claude-3-sonnet-20240229",
   "claude-3-haiku-20240307",
@@ -81,8 +76,9 @@ export const anthropicModels = [
 
 export type AnthropicModel = (typeof anthropicModels)[number];
 export const supportedModels = {
-  [ModelProvider.Anthropic]: anthropicModels,
-  [ModelProvider.OpenAI]: openAIModels,
+  [LLMAdapter.Anthropic]: anthropicModels,
+  [LLMAdapter.OpenAI]: openAIModels,
+  [LLMAdapter.Azure]: [],
 } as const;
 
 export type LLMFunctionCall = {
@@ -90,3 +86,26 @@ export type LLMFunctionCall = {
   description: string;
   parameters: z.ZodTypeAny; // this has to be a json schema for OpenAI
 };
+
+export const LLMApiKeySchema = z
+  .object({
+    id: z.string(),
+    projectId: z.string(),
+    createdAt: z.date(),
+    updatedAt: z.date(),
+    adapter: z.nativeEnum(LLMAdapter),
+    provider: z.string(),
+    displaySecretKey: z.string(),
+    secretKey: z.string(),
+    baseURL: z.string().nullable(),
+    customModels: z.array(z.string()),
+    withDefaultModels: z.boolean(),
+  })
+  // strict mode to prevent extra keys. Thorws error otherwise
+  // https://github.com/colinhacks/zod?tab=readme-ov-file#strict
+  .strict();
+
+export type LLMApiKey =
+  z.infer<typeof LLMApiKeySchema> extends LlmApiKeys
+    ? z.infer<typeof LLMApiKeySchema>
+    : never;

packages/shared/src/server/services/S3StorageService.ts

@@ -1,5 +1,4 @@
 import type { Readable } from "stream";
-import { env } from "@/src/env.mjs";
 import { GetObjectCommand, S3Client } from "@aws-sdk/client-s3";
 import { Upload } from "@aws-sdk/lib-storage";
 import { getSignedUrl } from "@aws-sdk/s3-request-presigner";
@@ -8,44 +7,52 @@ type UploadFile = {
   fileName: string;
   fileType: string;
   data: Readable | string;
+  expiresInSeconds: number;
 };
 
-class S3StorageService {
+export class S3StorageService {
   private client: S3Client;
+  private bucketName: string;
 
-  constructor() {
-    if (!S3StorageService.getIsS3StorageConfigured(env)) {
-      throw new Error("S3 bucket is not configured");
-    }
+  constructor(params: {
+    accessKeyId: string;
+    secretAccessKey: string;
+    bucketName: string;
+    endpoint: string;
+    region: string;
+  }) {
+    const { accessKeyId, secretAccessKey, bucketName, endpoint, region } =
+      params;
 
     this.client = new S3Client({
       credentials: {
-        accessKeyId: env.S3_ACCESS_KEY_ID,
-        secretAccessKey: env.S3_SECRET_ACCESS_KEY,
+        accessKeyId,
+        secretAccessKey,
       },
-      endpoint: env.S3_ENDPOINT,
-      region: env.S3_REGION,
+      endpoint,
+      region,
     });
+    this.bucketName = bucketName;
   }
 
   public async uploadFile({
     fileName,
     fileType,
     data,
+    expiresInSeconds,
   }: UploadFile): Promise<{ signedUrl: string }> {
     try {
       await new Upload({
         client: this.client,
         params: {
-          Bucket: env.S3_BUCKET_NAME,
+          Bucket: this.bucketName,
           Key: fileName,
           Body: data,
           ContentType: fileType,
         },
       }).done();
 
-      const expiresInOneHour = 60 * 60;
-      const signedUrl = await this.getSignedUrl(fileName, expiresInOneHour);
+      const signedUrl = await this.getSignedUrl(fileName, expiresInSeconds);
 
       return { signedUrl };
     } catch (err) {
@@ -57,42 +64,20 @@ class S3StorageService {
 
   private async getSignedUrl(
     fileName: string,
-    ttlSeconds: number,
+    ttlSeconds: number
   ): Promise<string> {
     try {
       return await getSignedUrl(
         this.client,
         new GetObjectCommand({
-          Bucket: env.S3_BUCKET_NAME,
+          Bucket: this.bucketName,
           Key: fileName,
           ResponseContentDisposition: `attachment; filename="${fileName}"`,
         }),
-        { expiresIn: ttlSeconds },
+        { expiresIn: ttlSeconds }
       );
     } catch (err) {
       throw Error("Failed to generate signed URL");
     }
   }
-
-  static getIsS3StorageConfigured(
-    currentEnv: Env,
-  ): currentEnv is S3ConfiguredEnv {
-    return Boolean(
-      currentEnv.S3_BUCKET_NAME &&
-        currentEnv.S3_ACCESS_KEY_ID &&
-        currentEnv.S3_SECRET_ACCESS_KEY &&
-        currentEnv.S3_ENDPOINT &&
-        currentEnv.S3_REGION,
-    );
-  }
 }
-
-export { S3StorageService };
-
-type Env = typeof env;
-type S3ConfiguredEnv = Env & {
-  S3_ACCESS_KEY_ID: string;
-  S3_SECRET_ACCESS_KEY: string;
-  S3_ENDPOINT: string;
-  S3_REGION: string;
-};

packages/shared/src/server/services/email/batchExportSuccess/BatchExportSuccessEmailTemplate.tsx

@@ -0,0 +1,83 @@
+import * as React from "react";
+import {
+  Body,
+  Button,
+  Container,
+  Head,
+  Heading,
+  Hr,
+  Html,
+  Img,
+  Preview,
+  Section,
+  Tailwind,
+  Text,
+} from "@react-email/components";
+
+interface BatchExportSuccessTemplateProps {
+  userName: string;
+  receiverEmail: string;
+  downloadLink: string;
+  batchExportName: string;
+  expiresInHours: number;
+}
+
+export const BatchExportSuccessEmailTemplate = ({
+  receiverEmail,
+  downloadLink,
+  userName,
+  batchExportName,
+  expiresInHours,
+}: BatchExportSuccessTemplateProps) => {
+  const previewText = `Download your data export from Langfuse`;
+
+  return (
+    <Html>
+      <Head />
+      <Preview>{previewText}</Preview>
+      <Tailwind>
+        <Body className="mx-auto my-auto bg-background font-sans">
+          <Container className="mx-auto my-10 w-[465px] rounded border border-solid border-[#eaeaea] p-5">
+            <Section className="mt-8">
+              <Img
+                src="https://static.langfuse.com/langfuse_logo_transactional_email.png"
+                width="40"
+                height="40"
+                alt="Langfuse"
+                className="mx-auto my-0"
+              />
+            </Section>
+            <Heading className="mx-0 my-[30px] p-0 text-center text-2xl font-normal text-black">
+              Your data export is ready
+            </Heading>
+            <Text className="text-sm leading-6 text-black">
+              Hello <strong>{userName}</strong>
+            </Text>
+            <Text className="text-sm leading-6 text-black">
+              Your data export{" "}
+              <span className="font-mono">{batchExportName}</span> is ready to
+              download. The download link will expire in{" "}
+              <strong>{expiresInHours} hours</strong>.
+            </Text>
+            <Section className="mb-4 mt-8 text-center">
+              <Button
+                className="rounded bg-black px-5 py-3 text-center text-xs font-semibold text-white no-underline"
+                href={downloadLink}
+              >
+                Download Export
+              </Button>
+            </Section>
+            <Hr className="mx-0 my-[26px] w-full border border-solid border-[#eaeaea]" />
+            <Text className="text-xs leading-6 text-[#666666]">
+              This email was intended for{" "}
+              <span className="text-black">{receiverEmail}</span>. If you were
+              not expecting this email, please delete it.
+            </Text>
+          </Container>
+        </Body>
+      </Tailwind>
+    </Html>
+  );
+};
+
+export default BatchExportSuccessEmailTemplate;

packages/shared/src/server/services/email/batchExportSuccess/sendBatchExportSuccessEmail.ts

@@ -0,0 +1,56 @@
+import { createTransport } from "nodemailer";
+import { parseConnectionUrl } from "nodemailer/lib/shared/index.js";
+import { render } from "@react-email/render";
+
+import { BatchExportSuccessEmailTemplate } from "./BatchExportSuccessEmailTemplate";
+
+type SendBatchExportSuccessParams = {
+  env: Partial<
+    Record<"EMAIL_FROM_ADDRESS" | "SMTP_CONNECTION_URL", string | undefined>
+  >;
+  receiverEmail: string;
+  downloadLink: string;
+  userName: string;
+  batchExportName: string;
+  expiresInHours: number;
+};
+
+export const sendBatchExportSuccessEmail = async ({
+  env,
+  receiverEmail,
+  downloadLink,
+  userName,
+  batchExportName,
+  expiresInHours,
+}: SendBatchExportSuccessParams) => {
+  if (!env.EMAIL_FROM_ADDRESS || !env.SMTP_CONNECTION_URL) {
+    console.error("Missing environment variables for sending email.");
+
+    return;
+  }
+
+  try {
+    const mailer = createTransport(parseConnectionUrl(env.SMTP_CONNECTION_URL));
+    const htmlTemplate = render(
+      BatchExportSuccessEmailTemplate({
+        receiverEmail,
+        downloadLink,
+        userName,
+        batchExportName,
+        expiresInHours,
+      })
+    );
+
+    await mailer.sendMail({
+      to: receiverEmail,
+      from: {
+        address: env.EMAIL_FROM_ADDRESS,
+        name: "Langfuse",
+      },
+      subject: "Your data export is ready",
+      html: htmlTemplate,
+    });
+  } catch (error) {
+    console.error(error);
+  }
+};

packages/shared/src/server/services/email/passwordReset/sendResetPasswordVerificationRequest.tsx

@@ -0,0 +1,88 @@
+/*
+ * To be used in the `sendVerificationRequest` function of the `email` provider of NextAuth.js.
+ */
+
+import * as React from "react";
+import {
+  Body,
+  Button,
+  Container,
+  Head,
+  Heading,
+  Html,
+  Img,
+  Preview,
+  Section,
+  Tailwind,
+  Text,
+} from "@react-email/components";
+import { createTransport } from "nodemailer";
+import { render } from "@react-email/render";
+import { type SendVerificationRequestParams } from "next-auth/providers/email";
+
+interface ResetPasswordTemplateProps {
+  url: string;
+}
+
+const ResetPasswordTemplate = ({ url }: ResetPasswordTemplateProps) => {
+  const previewText = "Reset your Langfuse password";
+  return (
+    <Html>
+      <Head />
+      <Preview>{previewText}</Preview>
+      <Tailwind>
+        <Body className="mx-auto my-auto bg-background font-sans">
+          <Container className="mx-auto my-10 w-[465px] rounded border border-solid border-[#eaeaea] p-5">
+            <Section className="mt-8">
+              <Img
+                src="https://static.langfuse.com/langfuse_logo_transactional_email.png"
+                width="40"
+                height="40"
+                alt="Langfuse"
+                className="mx-auto my-0"
+              />
+            </Section>
+            <Heading className="mx-0 my-[30px] p-0 text-center text-xl font-normal text-black">
+              Forgot your Langfuse password?
+              <br />
+              It happens to the best of us.
+            </Heading>
+            <Section className="mb-8 mt-8 text-center">
+              <Button
+                className="rounded bg-black px-5 py-3 text-center text-xs font-semibold text-white no-underline"
+                href={url}
+              >
+                Reset your password
+              </Button>
+            </Section>
+            <Text className="text-center text-xs leading-6 text-[#666666]">
+              If you do not want to change your password or didn&apos;t request
+              a reset, you can ignore and delete this email.
+            </Text>
+          </Container>
+        </Body>
+      </Tailwind>
+    </Html>
+  );
+};
+
+export async function sendResetPasswordVerificationRequest(
+  params: SendVerificationRequestParams
+) {
+  const { identifier, url, provider } = params;
+  const transport = createTransport(provider.server);
+  const htmlTemplate = render(<ResetPasswordTemplate url={url} />);
+  const result = await transport.sendMail({
+    to: identifier,
+    from: provider.from,
+    subject: `Forgot your password?`,
+    text: `To reset your Langfuse password, please confirm your email:\n${url}\n\nIf you do not want to change your password or didn't request a reset, you can ignore and delete this email.`,
+    html: htmlTemplate,
+  });
+  const failed = result.rejected.concat(result.pending).filter(Boolean);
+  if (failed.length) {
+    throw new Error(`Email(s) (${failed.join(", ")}) could not be sent`);
+  }
+}
+
+export default ResetPasswordTemplate;

packages/shared/src/server/services/email/projectInvitation/ProjectInvitationEmailTemplate.tsx

@@ -15,14 +15,13 @@ import {
   Text,
 } from "@react-email/components";
 
-import { env } from "@/src/env.mjs";
-
 interface ProjectInvitationTemplateProps {
   invitedByUsername: string;
   invitedByUserEmail: string;
   projectName: string;
-  recieverEmail: string;
+  receiverEmail: string;
   inviteLink: string;
+  emailFromAddress: string;
   langfuseCloudRegion?: string;
 }
 
@@ -30,8 +29,9 @@ export const ProjectInvitationTemplate = ({
   invitedByUsername,
   invitedByUserEmail,
   projectName,
-  recieverEmail,
+  receiverEmail,
   inviteLink,
+  emailFromAddress,
   langfuseCloudRegion,
 }: ProjectInvitationTemplateProps) => {
   const previewText = `Join ${invitedByUsername} on Langfuse`;
@@ -72,7 +72,7 @@ export const ProjectInvitationTemplate = ({
             </Text>
             <Section className="mb-4 mt-8 text-center">
               <Button
-                className="rounded bg-black px-5 py-3 text-center text-xs font-semibold text-background no-underline"
+                className="rounded bg-black px-5 py-3 text-center text-xs font-semibold text-white no-underline"
                 href={inviteLink}
               >
                 Accept Invitation
@@ -90,10 +90,10 @@ export const ProjectInvitationTemplate = ({
             <Hr className="mx-0 my-[26px] w-full border border-solid border-[#eaeaea]" />
             <Text className="text-xs leading-6 text-[#666666]">
               This invitation was intended for{" "}
-              <span className="text-black">{recieverEmail}</span>. This invite
+              <span className="text-black">{receiverEmail}</span>. This invite
               was sent from{" "}
-              <span className="text-black">{env.EMAIL_FROM_ADDRESS}</span>. If
-              you were not expecting this invitation, you can ignore this email.
+              <span className="text-black">{emailFromAddress}</span>. If you
+              were not expecting this invitation, you can ignore this email.
             </Text>
           </Container>
         </Body>