zen-artist/LLM.md

zen-artist — AI Knowledge Base

Project: zen-artist Organization: zenlm Repository: https://github.com/zenlm/zen-artist HuggingFace: https://huggingface.co/zenlm/zen-artist Last Updated: 2026-02-27

Overview

zen-artist is part of the Zen AI model family by Hanzo AI / Zen LM. It is the text-to-image product. Editing is a separate repo, zen-artist-edit.

This repo is scaffolding — configuration, examples, attribution. It ships no weights.

Base model

Tongyi-MAI/Z-Image — Apache 2.0. Single-stream diffusion transformer (ZImageTransformer2DModel, ~6.2B) with a Qwen3 text encoder (~4.0B), loaded via the diffusers ZImagePipeline. The upstream card is retained verbatim at base-model/README.md.

The base was previously declared as black-forest-labs/FLUX.1-dev, whose licence is non-commercial, while this repo declared Apache 2.0 — a contradiction, since a downstream cannot relicense an upstream. Retargeted to Z-Image so the declared licence and the actual base agree.

Rules for AI Assistants

1. ALWAYS update LLM.md with significant discoveries
2. NEVER commit model weights (*.safetensors, *.bin, *.gguf, *.pt)
3. NEVER commit symlinked files (CLAUDE.md, AGENTS.md, GEMINI.md, QWEN.md)
4. NEVER create random summary files — update THIS file only
5. Zen models are based on Qwen3 architecture
6. The base model must be permissively licensed (Apache 2.0 / MIT). Never point this repo at FLUX.1-dev or any other non-commercial base — the declared licence is Apache 2.0 and a downstream cannot widen an upstream licence.
7. src/examples/ is vendored Qwen-Image code. Its NOTICE entry is required by Apache 2.0 §4 — keep it. Never edit an upstream licence or notice.

Context

This file (LLM.md) is symlinked as CLAUDE.md, AGENTS.md, GEMINI.md, QWEN.md.

---

Part of the Zen AI family — Clarity Through Intelligence