FLUX 3 Image
FLUX 3 Image uses one endpoint for text-to-image generation, image edits and compositions from up to ten reference images. Bounding boxes in the prompt guide the placement or editing of individual elements.
Model details
The prompt identifies the desired change and the role of each reference. Layout boxes use top, left, bottom and right coordinates on a 0–1000 scale. The service also supports browser-based generation in the BFL playground.FLUX 3 Image official documentation
Release and access details
The 22 October update added custom dimensions and match-input sizing. Text-only generation can reach 16.8 megapixels, while requests with reference images are limited to 4 megapixels. Width and height override the resolution and aspect-ratio fields.FLUX release notes
Practical use
Practical guidance. Use reference roles and layout boxes when composition must be repeatable. Review both the edited region and the rest of the image. Disable grounding when the task should depend only on supplied material, then compare the result before changing other settings.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API, Provider product |
| Model identifiers | flux-3-image |
| Access terms | Proprietary hosted service; no public weights verified.FLUX 3 Image official documentation |
| Reference images | Up to tenFLUX 3 Image official documentation |
| Grounding | Web grounding is enabled by default and can be disabled.FLUX 3 Image official documentation |
Related models
- FLUX 3 Image Klein
Black Forest Labs · FLUX 3
Distilled 5.6-billion-parameter image transformer for generation, editing and compositions with up to ten references.
Launch date not confirmedAPI · Downloadable · Provider product - FLUX.2 [max]
Black Forest Labs · FLUX.2
High-capacity hosted image model for generation and reference-image editing.
Launch date not confirmedAPI · Provider product