[]([https://github.com/topics/video-editing](https://github.com/topics/text-guided-image-editing))
[](https://www.python.org/downloads/release/python-38/)
# FlowEdit
[Project](https://matankleiner.github.io/flowedit/) | [Arxiv](https://arxiv.org/abs/2412.08629) | [Proceedings](https://openaccess.thecvf.com/content/ICCV2025/html/Kulikov_FlowEdit_Inversion-Free_Text-Based_Editing_Using_Pre-Trained_Flow_Models_ICCV_2025_paper.html) | [Demo](https://huggingface.co/spaces/fallenshock/FlowEdit) | [ComfyUI](#comfyui-implementation-for-different-models) | [Data](https://github.com/fallenshock/FlowEdit/tree/main/Data)
#### [Recorded Talk](https://www.youtube.com/live/2fEDy-uTAII?si=_NRbANcqgX9wyvcI&t=17998)
### [ICCV 2025 Best Student Paper] Official Pytorch implementation of the paper: "FlowEdit: Inversion-Free Text-Based Editing Using Pre-Trained Flow Models"
## Installation
1. Clone the repository
2. Install the required dependencies using `pip install torch diffusers transformers accelerate sentencepiece protobuf`
* New version of diffusers may have compatibility issues, try install `diffusers==0.30.1`
* Tested with CUDA version 12.4 and diffusers 0.30.0
## Running examples
Run editing with Stable Diffusion 3: `python run_script.py --exp_yaml SD3_exp.yaml`
Run editing with Flux: `python run_script.py --exp_yaml FLUX_exp.yaml`
## Usage - your own examples
* Upload images to `example_images` folder.
* Create an edits file that specifies: (a) a path to the input image, (b) a source prompt, (c) target prompts, and (d) target codes. The target codes summarize the changes between the source and target prompts and will appear in the output filename.
See `edits.yaml` for example.
* Create an experiment file containing the hyperparamaters needed for running FlowEdit, such as `n_max`, `n_min`. This file also includes the path to the `edits.yaml` file
See `FLUX_exp.yaml` for FLUX usage example and `SD3_exp.yaml` for Stable Diffusion 3 usage example.
For a detailed discussion on the impact of different hyperparameters and the values we used, please refer to our paper.
Run `python run_script.py --exp_yaml `
## ComfyUI implementation for different models
* [FLUX](https://github.com/logtd/ComfyUI-Fluxtapoz)
* [HunyuanLoom](https://github.com/logtd/ComfyUI-HunyuanLoom)
Implemented by [logtd](https://x.com/logtdx/status/1869095838016012462?s=48&t=6Yj6BZKooDOmH_JWRWjtHg)
LTX-Video ComfyUI implementation can be found in LTX-Video [official repository](https://github.com/Lightricks/ComfyUI-LTXVideo/tree/master?tab=readme-ov-file#flow-edit).
## Community and Follow-Up Work
* [Training-Free-WAN-Editingš¤](https://github.com/KyujinHan/Awesome-Training-Free-WAN2.1-Editing), combines [WAN2.1](https://github.com/Wan-Video/Wan2.1) with FlowEdit to extend training-free to video editing. If you are interested in video editing, please feel free to take a look. Implemented by [Kyujinpy](https://github.com/KyujinHan).
* DNAEdit refines the Gaussian noise in the noise domain, improving image and video editing (NeurIPS 2025 Spotlight). [Project](https://xiechenxi99.github.io/DNAEdit/) | [Code](https://github.com/xiechenxi99/DNAEdit_code) | [Arxiv](https://arxiv.org/abs/2506.01430) | [Proceedings](https://neurips.cc/virtual/2025/loc/san-diego/poster/118684)
* FlowAlign add optimal control-based trajectory control to the inversion free process (ICLR 2026). [Code](https://github.com/FlowAlign/FlowAlign) | [Arxiv](https://arxiv.org/abs/2505.23145) | [Proceedings](https://openreview.net/forum?id=nyttIJfwW7)
* DynaEdit extened FlowEdit for dynmaic video editing. [Project](https://dynaedit.github.io/) | [Arxiv](https://arxiv.org/abs/2603.17989)
## License
This project is licensed under the [MIT License](LICENSE).
### Citation
If you use this code for your research, please cite our paper:
```
@inproceedings{kulikov2025flowedit,
title={Flowedit: Inversion-free text-based editing using pre-trained flow models},
author={Kulikov, Vladimir and Kleiner, Matan and Huberman-Spiegelglas, Inbar and Michaeli, Tomer},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={19721--19730},
year={2025}
}
```