[Feature]: Support for Black Forest Labs' flux-kontext-pro model in LiteLLM's /images/edits endpoint

[mvrodrig]

The Feature

Hi @krrishdholakia and @ishaan-jaff!

I'd like to request the addition of the new flux-kontext-pro model from Black Forest Labs as a supported provider for the /images/edits endpoint.

Black Forest Labs links:

• Flux Kontext
• Dashboard
• API Docs

Motivation, pitch

After testing, I’ve found that flux-kontext-pro delivers significantly better quality than gpt-image-1, especially in preserving the original faces in edited images (which is a major pain point with other models). It also provides faster inference times, making it a strong candidate for production use.

Adding support for this model would be highly valuable for users who need higher fidelity and speed in image editing tasks.

Related issues:

• #6066
• #6149

Thanks for considering this!

LiteLLM is hiring a founding backend engineer, are you interested in joining us and shipping to all our users?

No

Twitter / LinkedIn details

No response

[lanyuer]

+1

[lanyuer]

Support this proposal. As multimodal model applications become increasingly prevalent, would it be possible to add another mode—specifically, support for asynchronous model invocation (such as FLUX, Keling, etc.)—rather than adding more niche text model providers? In my own use case, I manage 4–5 providers and dozens of text/vision models via LiteLLM, but in practice I only use the OpenAI-compatible interface.