FLUX.1 Fill [pro]
FLUX.1 Fill [pro] is Black Forest Labs's dedicated inpainting model. It fills masked regions of images using text instructions while preserving the coherence, lighting, and style of the surrounding content.
import { experimental_generateImage as generateImage } from 'ai';
const result = await generateImage({ model: 'bfl/flux-pro-1.0-fill', prompt: 'A red balloon on a wooden table.'});What To Consider When Choosing a Provider
Zero Data Retention
AI Gateway does not currently support Zero Data Retention for this model. See the documentation for models that support ZDR.Authentication
AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
Inpainting workflows typically involve user-defined masks, so request shapes vary. Per-request cost tracking in AI Gateway helps you monitor spend on variable-complexity fill operations. See $0.05 for list pricing.
When to Use FLUX.1 Fill [pro]
Best For
Object removal and replacement:
Workflows where a specific element must be erased and replaced with contextually appropriate generated content
Background completion and outpainting:
Masking the edge of an existing image to generate new background content that extends the scene
E-commerce product cleanup:
Filling distracting elements, shadows, or artifacts cleanly in product photography
Generative object insertion:
Placing a new element into an image by masking the target region and describing the desired content in a text prompt
Consider Alternatives When
Instruction-based edits:
Kontext Pro or Kontext Max handle local edits like "change the color of the hat" better when no mask is used
Multi-reference consistency:
Kontext Pro or FLUX.2 Pro with multi-reference input are better suited for character consistency across scenes
Real-time interactive speed:
FLUX.2 Klein handles sub-second latency better than a fill-specialized model
Conclusion
FLUX.1 Fill [pro] brings a dedicated inpainting architecture to AI Gateway. For applications that need to remove objects, extend backgrounds, or insert new content into photographs and designs, Fill Pro is the purpose-built choice in the FLUX family.
FAQ
Inpainting takes an image plus a mask and generates pixels for the masked region only. Instruction-based editing takes an image plus natural language and changes the image without a pixel mask. Fill Pro is built for the first pattern; Kontext Pro and Max target the second.
See the pricing section on this page for today's rates. AI Gateway exposes each provider's pricing for FLUX.1 Fill [pro].
An input image, a mask defining the region to fill, and a text prompt describing the desired content for the masked area. This is the standard inpainting input format.
Yes. By placing the mask at the border or edge of an image, Fill Pro can generate new background content that extends the scene, a technique called outpainting.
Fill Pro is specialized for mask-based inpainting, generating content for a defined region. The Kontext models (Pro and Max) are specialized for text-instruction-based editing that understands the full image context and makes modifications described in natural language, without requiring an explicit pixel mask.
No. Fill Pro is a dedicated image generation model, not a language model or multimodal LLM. Use generateImage from the AI SDK; it processes image and mask inputs to produce image outputs.
Fill Pro is purpose-built for inpainting. For text-to-image generation without an existing image, use FLUX.1 Pro, FLUX.2 Pro, or other members of the FLUX family.
