HunyuanImage-2.1
HunyuanImage-2.1 is an earlier 17B diffusion image generator. It uses a base generation stage and a refiner to produce 2K images.
Model design and access
A multimodal language encoder supports prompt alignment, while a multilingual character-aware encoder supports text rendering. The card warns that 1K generation produces artifacts. Its supported sizes vary with aspect ratio; 2K does not mean every output is square.HunyuanImage-2.1 official model documentation
Practical use
Practical guidance. Use one of the documented output sizes. Review any later resizing separately from model generation, especially for small text. Keep the refiner enabled or disabled consistently when comparing results.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| License | Tencent Hunyuan Community License; custom terms applyHunyuanImage-2.1 official model documentation |
| Resolution | 2K image generation; 1K is not supportedHunyuanImage-2.1 official model documentation |
| Compression | 32 × 32 spatial VAE compressionHunyuanImage-2.1 official model documentation |
Related models
- HunyuanImage-3.0
Tencent · HunyuanImage
An autoregressive text-to-image model with 80B total and 13B active parameters.
Launch date not confirmedDownloadable