Difflet
Runs diffusion transformer image and video models via CLI, Python API, and server.
Tested on Linux
Use it
Linux x86_64
You need first
- Runtimes:
python>=3.10 - A GPU is optional
- Keys or accounts: Hugging Face account for gated repos such as FLUX.1-dev
clone Difflet repo
git clone https://github.com/ai-decentralized/Difflet.gitpip install torch cu128 from CUDA index
pip install --index-url https://download.pytorch.org/whl/cu128 torch==2.11.0pip install -e .[cuda] in repo
cd Difflet && pip install -e ".[cuda]"
Check that it works
difflet --helpThe output should contain “usage: difflet”.
First thing to try
difflet cache lsRun difflet cache ls to list the models in your local cache.
What it does
- Run diffusion transformers on different acceleratorsNot tested
- Generate an image or video with one commandNot tested
- Serve models over an OpenAI-compatible APINot tested
- Generate from Python with a diffusers-style APINot shown working
How far pickfit got
Linux x86_64
- Installs
- Starts
- 0 of 4 features work
- Works offline
Good for
- generate a 1024x1024 image from a text prompt
- generate a video from a text prompt
- serve image and video models over OpenAI-compatible HTTP
- generate text-to-video with audio via staged CLI
How pickfit tested it
Each run starts from a clean machine and follows the steps above. How pickfit verifies
Linux x86_64
Its main function works
- Replayed from scratch
- Works offline
- Checked on 1 Nov 2026
- Version 0.0.0.dev0