FLUX.1 Kontext Dev
Edits one photo from a written instruction using Black Forest Labs' FLUX.1 Kontext [dev]. Ask it to remove a parked car or add rain, and the rest of the scene carries over, redrawn as a 1024 by 1024 square.
Overview
This page runs FLUX.1 Kontext [dev], the 12B-parameter rectified flow transformer Black Forest Labs released for editing images from text instructions. Upload one picture and write the change, for example "Remove the car." In the default sample the silver sports car disappears and the street around it is left intact. Black Forest Labs says the model handles local and global edits and keeps visual drift low over successive edits, so a result can go back in for the next change.
Choosing a variant: this page takes a single image. FLUX.1 Kontext Dev 2 Images adds a second upload for placing something from another photo into the scene, and FLUX.1 Kontext Dev 3 Images builds one picture from three. Pick this one when everything the edit needs is already in the photo.
What you can build
- Color variants of a product image: "Change the color of the car to purple." turns a silver pickup truck purple.
- Weather changes. "Add rain effect to the photo." draws rain streaks over a beach.
- Subject swaps, such as a sunflower replaced by a white rose.
Inputs
- One image, uploaded or given as a link.
- A prompt describing the change (required).
- Advanced settings: inference steps from 1 to 500 (28 by default), guidance scale from 1 to 20 (2.5 by default, the value in Black Forest Labs' reference example), 1 to 8 outputs, and a seed.
Outputs
Every sample result is a 1024 by 1024 square, even the car edit, whose source is a 1536 by 2048 portrait. Size and aspect ratio have no setting.
Limitations
- There is no mask input. The prompt alone tells the model where to work, so name the object precisely when the photo holds several similar ones.
Safety & compliance
The weights are published under the FLUX.1 [dev] Non-Commercial License, and the model card says generated outputs may be used for personal, scientific and commercial purposes as described there. Black Forest Labs' usage terms prohibit unlawful content and the use of outputs for unlawful, defamatory or abusive purposes. If the photo shows a real person, get their consent first.
Example prompts
Great starting points for flux_kontext_dev.
API quick start
Run flux_kontext_dev with a single API call.
{
"prompt": "Remove the car.",
"inputImage": "https://your-cdn.com/input.png",
"steps": 28,
"scale": 2.5
}curl -X POST "https://api.wiro.ai/v1/Run/wiro/flux-kontext-dev" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"prompt": "Remove the car.",
"inputImage": "https://your-cdn.com/input.png",
"steps": 28,
"scale": 2.5
}
JSONOther models you might like
Explore similar image generation models.



