HswAI2026/JuZhou-V1

Health v2 · maintenance only

Flags: no_releases young no_license

Adoption not part of the score

1097 stars · 194 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

JuZhou 1.0 is an ultra-lightweight 0.387B-parameter text-to-image foundation model designed for fully offline, on-device execution on mobile hardware. It was trained entirely on China-developed Sugon K100 AI accelerators and features native Chinese semantic alignment from 9M curated Chinese image-text pairs.

Use cases

• generate images from text prompts offline on a phone
• run a lightweight text-to-image model on mobile devices
• generate images from Chinese text prompts without translation
• privacy-preserving local image generation without cloud APIs
• deploy a compact diffusion model on edge hardware
• compare small T2I model quality against SDXL and SD3

When to choose

• you need offline, on-device image generation on mobile
• your prompts are in Chinese and you want native semantic alignment
• you need a small model that fits edge hardware constraints
• privacy rules out sending prompts to a cloud image API

When to avoid

• you need the highest possible image fidelity from large frontier models
• you require English-first prompt understanding
• you need fine-grained control like inpainting or ControlNet workflows
• you depend on NVIDIA CUDA tooling for training or inference

Facets

library · maturity active

machine-learning image-processing llm-inference artificial-intelligence image-processing mobile-development cross-platform text-to-image diffusion-model on-device-inference chinese-language edge-ai stable-diffusion offline android mobile

Member repositories

For agents

markdown · JSON · MCP: product_card(name="HswAI2026/JuZhou-V1")