Tongyi-MAI/Z-Image

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

11944 stars · 813 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Z-Image is a 6B-parameter text-to-image generation foundation model family built on a single-stream diffusion transformer, with a distilled Turbo variant that runs in 8 function evaluations and fits in 16GB VRAM. The repository provides inference code, checkpoints, and demos for generating photorealistic images with bilingual (English/Chinese) text rendering.

Use cases

• generate images from text prompts
• run fast text-to-image generation on consumer GPUs
• render text inside generated images in English and Chinese
• generate photorealistic images with a diffusion model
• run image generation locally with 16GB VRAM
• distill or fine-tune an image generation foundation model

When to choose

• you need high-quality text-to-image generation with low inference latency
• you want bilingual text rendering in generated images
• you have a consumer GPU with 16GB VRAM and want local image generation
• you want an Apache-2.0 licensed image generation model you can self-host

When to avoid

• you need video or audio generation rather than images
• you have no GPU and cannot run diffusion inference locally
• you need a hosted API without managing model weights yourself

Facets

library · maturity active

image-processing machine-learning deep-learning llm-inference image-processing deep-learning artificial-intelligence python cross-platform text-to-image diffusion-transformer image-generation diffusion-model bilingual-text-rendering model-weights gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="Tongyi-MAI/Z-Image")