HiDream-O1-Image_clear_v1

• Guide (External site): English | Japanese

• Hugging Face

A finetuned HiDream-O1-Image model for clear outputs.

Compare to HiDream-O1-Image (Original)

HiDream-O1-Image_clear_v1 | HiDream-O1-Image

Compared to the original, this model generates images with higher contrast, and richer colors.

ComfyUI workflow

Note: Due to Civitai's restrictions, these images do not contain the workflow metadata. Please download the workflow including the metadata from Hugging Face.

Text to Image

Image to Image

Image Edit

To use these workflow, please install ComfyUI-uit-hidream-o1 custom node.

Model Format and Quality

Generally, model size and image quality are proportional.

• FP32 > FP16 ≒ BF16 > MXFP8 > FP8_scaled > NVFP4 > FP8

Generation time varies greatly depending on whether the model can be loaded into VRAM, and the supported formats are determined by the GPU generation.

MXFP8 and NVFP4 are the latest standards designed for the RTX 5000 series. On unsupported GPUs, however, fallback mechanisms will actually slow down processing, so their use is not highly recommended.

For a comparison of image quality and generation time by format, please see this page.

Recommended Settings

• Sampler: euler

• Scheduler: normal

• Steps: 12 (8-24)

• CFG scale: 4 (3-5)

• noise_scale: 6.5 (6.2-7.5)

Tips

• HiDream-O1-Image and UiT architecture: English | Japanese

• FP / BF Format Image Quality and Generation Time: English | Japanese

• Image Generative AI model License: English | Japanese

License

• MIT