FLUX.1 Kontext Dev 2 Images
Places something from one photo into another with FLUX.1 Kontext [dev]. Upload a scene and a second picture, such as a toy dog or a stone texture, and say in the prompt where it belongs.
Overview
Two uploads and one instruction go into FLUX.1 Kontext [dev], Black Forest Labs' 12B-parameter image editing model. Usually one is the scene to keep and the other holds what to bring in. The default pair is a living room and a felt toy dog with the prompt "Add the dog as if it's lying on the couch." The result keeps the sofa, rug and shelves and puts the dog on the cushions. Black Forest Labs lists character, style and object reference among the model's features.
Choosing a variant: for a change that needs nothing outside the photo, use FLUX.1 Kontext Dev. FLUX.1 Kontext Dev 3 Images assembles a scene from three sources. Pick this page when exactly one outside element has to join an existing photo.
What you can build
- Pet or product placement, as in the dog sample.
- Interior previews: one sample puts stonework from a texture photo on the wall behind a bed and leaves the rest of the bedroom alone.
- Logo mockups on shirts, which two of the page's example prompts ask for.
Inputs
- Two images, both required, in JPG, JPEG, PNG, GIF, WEBP or HEIC.
- A prompt (required). The wall example names each picture's role first: "The first image is the room. The second image is the pattern."
- Advanced settings: 1 to 500 steps (default 28), guidance scale 1 to 20 (default 2.5), up to 8 outputs and a seed.
Outputs
All three results shown are 1024 by 1024. A wide family photo, with its black dog swapped for the Labrador from the second upload, came back reframed: the people are shown full length, with floor drawn in below the original crop.
Limitations
- Placement and scale come from the wording alone, as in "Print the logo in small size at the center of the woman's shirt."
Safety & compliance
The FLUX.1 [dev] Non-Commercial License covers this model, and the card notes that generated images may serve personal, scientific and commercial purposes within its terms. Black Forest Labs' policies rule out unlawful content as well as unlawful, defamatory or abusive use of results. Anyone recognizable in either upload should have agreed to the edit.
Example prompts
Great starting points for flux_kontext_dev_with_2_image.
API quick start
Run flux_kontext_dev_with_2_image with a single API call.
{
"prompt": "Add the dog as if it's lying on the couch.",
"inputImage": "https://your-cdn.com/input.png",
"inputImage2": "https://your-cdn.com/input.png",
"steps": 28
}curl -X POST "https://api.wiro.ai/v1/Run/wiro/flux-kontext-dev-with-2-image" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"prompt": "Add the dog as if it's lying on the couch.",
"inputImage": "https://your-cdn.com/input.png",
"inputImage2": "https://your-cdn.com/input.png",
"steps": 28
}
JSONOther models you might like
Explore similar image generation models.



