Black Forest Labs' FLUX.1 Kontext uses an existing image together with written instructions to generate an edited version. Its announcement describes targeted changes, reference-based styling, and successive edits intended to preserve recognizable characters and objects. This gives creators a way to refine an image through instructions while retaining visual context windowThe maximum amount of text a model can consider at once — its working memory for the current conversation or task.Full definition → from the previous version.
Continuity has limits. The company reports that long sequences of edits can introduce visible artifacts and degrade image quality. Kontext therefore supports iterative revision, but preserving a character or composition across edits still requires checking the resulting images.
Terms in this piece · Glossary
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
