FLUX.1 Kontext Max
FLUX.1 Kontext Max is Black Forest Labs's experimental maximum-fidelity editing tier. It prioritizes typography accuracy, precise prompt adherence, and edit consistency for tasks where exact instruction fidelity is the primary requirement.
import { experimental_generateImage as generateImage } from 'ai';
const result = await generateImage({ model: 'bfl/flux-kontext-max', prompt: 'A red balloon on a wooden table.'});What To Consider When Choosing a Provider
Zero Data Retention
AI Gateway does not currently support Zero Data Retention for this model. See the documentation for models that support ZDR.Authentication
AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
Black Forest Labs labels Kontext Max as experimental. It's the right model when you need the highest Kontext tier. If you prioritize throughput and iteration speed over maximum fidelity, evaluate Kontext Pro first. See $0.08 for current pricing.
When to Use FLUX.1 Kontext Max
Best For
Typography-critical editing:
Changing text content, correcting signs or labels, or adding legible on-image text while preserving surrounding context
High-precision prompt following:
Detailed instructions specifying exact colors, precise layouts, or specific visual attributes must execute as literally as possible
Higher-fidelity edits:
Production workflows where Kontext Pro's edits don't follow the prompt closely enough
Character and identity preservation:
Complex editing sequences where consistency across turns is a hard requirement
Consider Alternatives When
High-volume iteration:
Kontext Pro targets fast-iteration editing where throughput and speed matter more than maximum precision
Pure text-to-image generation:
FLUX.2 Pro or FLUX.2 Max offer better raw generation quality when no reference image is used
Real-time agent editing:
FLUX.2 Klein targets sub-second latency inside an agent loop
Conclusion
FLUX.1 Kontext Max sits at the top of the Kontext family for prompt adherence and on-image text accuracy. When Kontext Pro's output doesn't match the editing instruction closely enough, or when text rendering accuracy is non-negotiable, Kontext Max is the appropriate next step.
FAQ
Black Forest Labs positions Kontext Max above Kontext Pro for prompt adherence and on-image text, at the same inference speed. Use Max when an edit instruction or on-image text must match the prompt more literally than Kontext Pro delivers.
Rendering legible, accurate text within images is difficult for many generative models. Kontext Max targets on-image text edits: it can change text in a sign, label, or UI element while preserving surrounding context. It scores higher on those tasks than Kontext Pro.
Yes. Kontext Max shares the multi-turn iterative editing capability of the Kontext suite. Each output can serve as the reference image for the next editing step. That keeps character and style aligned across many sequential changes.
Choose Max when the exact language of the edit instruction must be followed with maximum fidelity, or when text within images must be rendered accurately. Choose Pro when speed and iteration throughput are the primary concerns and moderate prompt adherence is sufficient.
Black Forest Labs labels Kontext Max as experimental, meaning it's still being refined. The model is fully accessible through AI Gateway. Evaluate it on your specific editing tasks to confirm it meets your quality and consistency requirements before committing to a production integration.
Kontext models achieve fast inference speeds relative to many competing models, and Kontext Max's improvements come "without compromise on speed."
Current pricing is shown on this page. AI Gateway routes across providers, and rates may vary by provider.
