mirror of
https://github.com/hanzo-apps/hanzo.ai.git
synced 2026-10-30 14:58:08 +00:00
zooqueen
75c4f1f641
The CLI lines on the product pages were written from the product name -- hanzo dns zones, hanzo kms keys -- and several were not commands the binary has. So the command moves into lib/product.ts, which assembles the rest from what is already stated: the catalog names the product and its route, pages.json names what the catalog does not carry, and products-metadata answers for docs, where null means docs publishes nothing and the link drops. Every CLI line in that table was run against the binary. A product absent from it has no command and its page prints none; nothing composes a verb out of a name. Both blocks that draw it import the Product type rather than declaring their own, which leaves the table the only place a command can be written. The catalog also moves out of package.json into pnpm-workspace.yaml, where pnpm reads it, so the versions are stated once.
80 lines
3.8 KiB
TypeScript
80 lines
3.8 KiB
TypeScript
'use client'
|
|
|
|
import { ArrowRight, Boxes, Cpu, Gauge, Layers, Server, Shield, Zap } from 'lucide-react'
|
|
import { ProductLanding } from '@/components/product/ProductLanding'
|
|
import { ProductFooter } from '@/components/products/ProductFooter'
|
|
|
|
/**
|
|
* Hanzo Engine, through the ONE product kit.
|
|
*
|
|
* The largest cluster left: TWELVE pages — engine, gateway, guard, ingress,
|
|
* jin, metrics, o11y, operator, sentinel, skills, telemetry, visor — at 85%+
|
|
* identical className sets, each a hero, a six-card grid and a call to action
|
|
* written out again. All twelve pass one typed props object to
|
|
* `ProductLanding`. See LLM.md, "Tailwind is NOT gone".
|
|
*
|
|
* EVERY WORD IS THE ONE THAT WAS HERE. ProductFooter came out of the final
|
|
* card's `max-w-3xl` wrapper, where it had been rendering narrower than on
|
|
* every other product page.
|
|
*/
|
|
|
|
export default function HanzoEnginePage() {
|
|
return (
|
|
<>
|
|
<ProductLanding
|
|
badge={"hanzoai/engine"}
|
|
badgeIcon={Cpu}
|
|
title={"Hanzo Engine"}
|
|
tagline={"The inference engine, in one binary"}
|
|
lede={"Name a model on Hugging Face and it runs. Engine quantizes it for the hardware it finds, then serves an HTTP API and a web chat from the same process — text, vision, audio, speech, image and embeddings, without a second stack underneath."}
|
|
ctas={[
|
|
{ label: "View on GitHub", href: "https://github.com/hanzoai" },
|
|
]}
|
|
what={{
|
|
title: "One process, laptop to GPU cluster",
|
|
sub: "Chat, an API server, and a benchmark are three commands on the same binary. There is no Python interpreter under any of them.",
|
|
pillars: [
|
|
{
|
|
icon: Zap,
|
|
title: "run, serve, bench",
|
|
body: "One binary, three verbs. Chat with a model, put it behind an HTTP API, or measure it — the thing you benchmarked is the thing you shipped.",
|
|
},
|
|
{
|
|
icon: Gauge,
|
|
title: "Built to serve, not just to run",
|
|
body: "Paged attention and continuous batching keep the card busy across concurrent requests. A disk-first KV cache carries a session across a restart, so an agent that reuses a long prefix does not pay for it twice.",
|
|
},
|
|
{
|
|
icon: Layers,
|
|
title: "More than text",
|
|
body: "Vision, audio, speech, image and embeddings go through the same modality pipeline as generation, in the same process. A multimodal app is one deployment, not four.",
|
|
},
|
|
{
|
|
icon: Boxes,
|
|
title: "It quantizes on the way in",
|
|
body: "In-situ quantization of any Hugging Face model, plus GGUF from 2 to 8 bits, GPTQ, AWQ, HQQ, FP8 and bitsandbytes. Ask for a level and it takes a published prebuilt when one exists, or makes one when it doesn't.",
|
|
},
|
|
{
|
|
icon: Server,
|
|
title: "The silicon you have",
|
|
body: "CUDA with FlashAttention and multi-GPU tensor parallelism, Metal on Apple Silicon, or plain CPU. Same engine and same commands on each, and a doctor command that reports what it found.",
|
|
},
|
|
{
|
|
icon: Shield,
|
|
title: "Tools, and a UI to watch them",
|
|
body: "It can call MCP servers over a process, HTTP or a WebSocket. The built-in UI at /ui shows reasoning, code execution and plots inline, and edits branch a conversation with its own state. Pass --no-ui and it is a plain API server.",
|
|
},
|
|
],
|
|
}}
|
|
finalCta={{
|
|
title: "Get started with Hanzo Engine",
|
|
buttons: [
|
|
{ label: "View on GitHub", href: "https://github.com/hanzoai" },
|
|
],
|
|
}}
|
|
/>
|
|
<ProductFooter slug={"engine"} name={"Engine"} />
|
|
</>
|
|
)
|
|
}
|