HiDream-I1 Native Support in ComfyUI!
Hi Comfy community! We'd like to share the new native capabilities for the HiDream-I1 model.
About HiDream-I1
HiDream-I1 is a text-to-image model officially open-sourced by HiDream-ai on April 7, 2025. The model has 17B parameters and is released under the MIT license, supporting personal projects, scientific research, and commercial use. It currently performs excellently in multiple benchmark tests.
HiDream-ai originally provides 3 versions of the models:
HiDream-I1-Full: Full version, 50 inference steps 🤗 HiDream-I1-Full
HiDream-I1-Dev: Distilled version, 28 inference steps 🤗 HiDream-I1-Dev
HiDream-I1-Fast: Distilled version,16 inference steps 🤗 HiDream-I1-Fast
Get Started
Update ComfyUI or ComfyUI Desktop to the latest
Go to “Template” and select “HiDream” workflows, or download workflows below
The workflow will prompt you to download the model
Start generating!
For manual file downloads, check our documentation for details.
1. HiDream-I1 Full Version Workflow
Example Outputs:
2. HiDream-I1 Dev Version Workflow
Example Outputs:
3. HiDream-I1 Fast Version Workflow
For more tips and details, please check our documentation.
Enjoy creating!
Hey guys, thank you very much for implementing HiDream in ComfyUI.
I am one of the first testers of this model on ComfyUI with the custom node "comfyui_HiDream-Sampler". This is another big step for the open source generative AI community after the release of FLUX. I'm so excited to explore this new models capabilities.
I'd like to drop the models URL here (in safetensor file format) for those who are looking for the fp8 and bf16 version of HiDream like me:
https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/tree/main/split_files/diffusion_models
Appreciate ComfyUI and your continuous updates and adding new models for us mere mortals to use :)