Black Forest Labs drops FLUX.1 Kontext for in-context image editing
Black Forest Labs just rolled out FLUX.1 Kontext, a new suite of models that generate and edit images using text and visual references side-by-side. It handles local tweaks and character consistency without complex workflows, accompanied by an open-weight 12B dev model on the BFL Playground.

FLUX.1 Kontext is a new generative flow-matching model suite from Black Forest Labs built to handle both text-to-image creation and targeted editing in one place.
Instead of fiddling with convoluted node graphs or training custom LoRAs just to change an outfit, Kontext lets you pass existing images alongside text prompts. The model parses the visual context directly and makes clean edits on the fly.
What you can do with it
The standout feature here is in-context editing that actually keeps character consistency intact across multiple turns.
You can take a portrait, ask the model to wipe away an unwanted object, move the subject onto the streets of Freiburg, or drop heavy snow into the background. The core subject stays recognizable while the rest of the scene transforms.
Beyond keeping faces and objects steady, Kontext handles targeted regional tweaks, style transfers pulled from a reference image, and sharp typography rendering without blowing out the rest of the composition.
Three models and an open 12B weight
Black Forest Labs rolled the lineup out across three tiers.
Kontext [max] sits at the top for raw performance, dialing up prompt following, typography rendering, and strict editing consistency at high throughput.
Kontext [pro] is aimed at fast, iterative creative workflows, packing full-scene generation and surgical local tweaks into a unified setup.
The one makers should grab immediately is FLUX.1 Kontext [dev]. It is an open-weight, 12B parameter model built to deliver proprietary-grade editing directly on consumer-grade hardware.
You can jump right in on the new web-based BFL Playground, or grab the dev weights to wire it into your local pipeline.
Why it matters
Image generation has been good enough for concept art for a while, but real production work dies the moment you need consistent character continuity or simple local tweaks. Kontext baking in-context editing directly into base model weights changes that calculus entirely.
For creators and indie developers tired of babysitting brittle ComfyUI workflows, the 12B dev model looks like a must-try. It turns multi-turn editing into natural conversation, bringing open models much closer to real commercial utility.
Sources (2)
- x.com(opens in a new tab)Originalx.com
- bfl.ai(opens in a new tab)Reportbfl.ai