AI GitHub
FLUX reference code for image generation
Explore Black Forest Labs' official inference code and model references.
Key Takeaways
- The repository documents model variants and inference entry points.
- Choose the checkpoint and precision that fit your intended generation workflow.
- Use a documented variant and compare a simple object prompt with one that needs precise composition or text.
- Model variants have distinct usage terms and hardware requirements.
How it works
The repository documents model variants and inference entry points. Each variant's weights and implementation choices determine its available workflows and resource needs.
What you need
Choose the checkpoint and precision that fit your intended generation workflow. Review its model-specific usage instructions before combining it with tools from another variant.
What to try first
Use a documented variant and compare a simple object prompt with one that needs precise composition or text. Keep dimensions and seed fixed when changing settings.
Include a realistic negative example for the composition you care about. A prompt that works on a flagship model variant may behave differently with another checkpoint.
Where to look for failures
Model variants have distinct usage terms and hardware requirements. Read the license associated with the exact checkpoint you select.
Related Franklin pages
Match the pipeline and device to a specific model variant. Compare a controlled prompt on known assets before testing a new checkpoint, conditioning input or export.
- Convert supported diffusion models and run them through Apple's Core ML stack. Apple Stable Diffusion for Core ML deployment
- Use modular diffusion pipelines for supported image, video and audio models. Diffusers for building generation pipelines in Python
Franklin’s image tests give you examples of outputs to compare across visual models. Franklin’s image tests
Project documentation
Try it for your workflow
Explore the project’s documentation for setup and usage, or save this page to a collection for later.
Read the project documentation ↗