Image model · from 6 credits
Generate images with Flux Kontext
Flux Kontext never starts from scratch: it requires an input image and transforms it on instruction. Swapping a background, fixing an object, restyling a scene while keeping its structure, that is its exclusive ground. The constraint is its strength: because it leans on your image, it preserves what pure generation models recompose at random. In the canvas you wire it behind an existing image node, never alone, and describe only what should change.
Capabilities
- Edits existing images and takes visual references
A model that never starts from nothing
Flux Kontext is not a generator like the others: it requires an input image and works on it. You describe what should change, not the whole scene, and it applies the edit while preserving what you did not mention. Swapping a background, removing an intrusive object, changing a colour, moving a photo into another graphic style are its typical uses.
That constraint is its strength. Ask a pure generator for the same scene with one variation and it recomposes everything, so the face, the framing and the details shift along the way. Kontext leans on your image, so what you wanted to keep stays put.
Writing an edit instruction
The rule is to describe the change, not the finished result. Replace the background with a pale concrete wall works better than redescribing the subject, the light and the framing in the hope that the model finds the original again. Naming explicitly what must stay untouched helps too, especially on faces and on text already present in the image.
Edits chain well. Rather than one instruction asking for four changes at once, prefer four successive steps, each one checkable. On the canvas that means wiring a second Kontext node behind the first: each step starts from the previous image.
Where it fits in a chain
Kontext always sits behind something: an image generated by another model, a photo imported into a Media node, or the output of a retouch. It does not work alone, and the canvas makes that plain by refusing to run a node with no image input.
In practice the most common chain runs like this: generate a visual with a Flux 2 model, fix what is wrong with Kontext, then animate the result in a Video node. For local corrections by brush or lasso the Retouch node is the better tool; Kontext works on the whole image, from one sentence.
Flux Kontext next to its closest siblings
Figures pulled from the same registry as billing: when a price or a cap changes, this table changes with it.
| Flux Kontext | Flux Pro 1.1 | Nano Banana | |
|---|---|---|---|
| Base price | 6 credits / image | 6 credits / image | 6 credits / image |
| Image editing | Yes | No | Yes |
What Flux Kontext produces
Renders made on Imaginode with this model, untouched.
How to use it on Imaginode
- 1Create a project and add an Image node from the toolbar.
- 2Pick Flux Kontext in the selector: the exact price shows on the Generate button.
- 3Write your prompt (the magic wand can enrich it for 1 credit) and generate. Technical failures are refunded automatically.
Frequently asked questions
Can Flux Kontext be used without a starting image?
No. It is an editing model: it needs an input image, either generated in the canvas or imported from your files. Without one, the node refuses to run.
Does it keep a person's face intact?
That is its design goal, and it does well when the instruction is about something other than the face. State in the instruction what must stay unchanged, and work through small successive edits rather than one large request.
How does it differ from the Retouch node?
The Retouch node works by hand, with brush, lasso and layers, on an area you outline. Kontext works on the whole image from a sentence. The two complement each other and wire one behind the other in the same project.
How much does an image generated with Flux Kontext cost?
6 credits per image, about $0.07 (1 credit = $0.012). The exact cost shows on the Generate button before you click, and any failed generation is refunded automatically within seconds.
Can Flux Kontext edit an existing image?
Yes: Flux Kontext accepts an input image. On the Imaginode canvas, wire an image node or an uploaded photo, describe the change, and the model starts from your visual instead of generating from scratch.
Can I try Flux Kontext for free?
Yes: signing up grants trial credits, no credit card required. Enough to test Flux Kontext in real conditions before deciding on a plan or a top-up.
Do I need to install anything to use Flux Kontext?
No: Flux Kontext runs server-side, in the browser, mobile included. Imaginode's node canvas wires its renders to the other models (references, camera, styles) with nothing to install.
Try Flux Kontext on a real canvas
105 models, one account, credits that work everywhere. The price always shows before you generate.
Get started
Flux Schnell
Seedream 5 Lite