Flux Kontext: change one locked still (4K, 4cr)
Flux Kontext is an edit. Point at the object. A new text-to-image throws away the still you approved.
Flux Kontext is an edit. Point at the object. A new text-to-image throws away the still you approved.
Flux Kontext is Flux's edit-image row: image required, 4K ceiling, 4 credits, a discounted catalog line. The description is "Context-aware generation with coherent compositions." Features listed: Context-aware, Coherent, Smart. Context-aware is not a synonym for "make me a new poster." The context is the still you already locked. The AI photo editor is the surface.
Context is the locked frame
You have to put an image in. The record lists a max of one reference. That one file is the composition you are keeping. The prompt names the object that changes. Coherent composition is the promise: the rest of the frame should still read as the same picture.
A prompt that restages the room is a text-to-image brief wearing an edit model's name. You will get a new still. You will not get the still the client approved. Image-to-image only holds when the instruction is smaller than the frame.
4K, 4 credits, image in
Four credits is the catalog price. Max output resolution is 4K, with SD and HD also listed. Aspect options are 1:1, 9:16, 16:9, 2:3, 3:2, 3:4, and 4:3. Styles listed are Realistic, Artistic, and Anime. None of those knobs mean "start over." They mean finish the plate.
Use this when the layout is signed and you still owe a SKU swap, a background cleanup, or a label fix. Do not use it as a cheaper Flux stills engine. If there is no approved plate, you are on a text-to-image row. The Flux hub is the rest of the family. This page is only Kontext: one still, one object, 4K, 4 credits.
Point, don't re-roll
The failure mode is "same vibe, different bottle" as a full-frame rewrite. Name the bottle. Name the change. Look. Then stop or do a second 4-credit pass. Stacking five edits in one prompt is how context-aware averaging eats the lock.
The image-to-image editing workflow is the pack-scale version. Flux Kontext is the single-still version. When the still is actually locked, motion is a different generate. Do not buy video to hide an unlocked label.
FAQ
Can I run Flux Kontext from text only?
No. The row requires an image. Text-only is a different Flux (or other) stills row, and it will not preserve the approved plate.
How many images can I feed?
The record lists a max of one reference. That one still is the lock. Extra concepts belong in the instruction, not in a pile of uploads this row will not take.
What does "context-aware" mean here?
The catalog description is context-aware generation with coherent compositions. In practice: the locked frame is the context. Point at the object. Do not restage.
Is 4 credits a 4K-only price?
The catalog price for this generate is 4 credits. Max output resolution is 4K. The row is discounted in the catalog. Neither fact is a reason to throw the still away.