Flux Pro V1 API

Fill
Fill Finetuned

Endpoint: POST https://fal.run/fal-ai/flux-pro/v1/fill Endpoint ID: fal-ai/flux-pro/v1/fill

Try it in the Playground

Run this model interactively with your own prompts.

Quick Start

import fal_client

def on_queue_update(update):
    if isinstance(update, fal_client.InProgress):
        for log in update.logs:
           print(log["message"])

result = fal_client.subscribe(
    "fal-ai/flux-pro/v1/fill",
    arguments={
        "prompt": "A knight in shining armour holding a greatshield with \"FAL\" on it",
        "image_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/knight.jpeg",
        "mask_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/mask_knight.jpeg"
    },
    with_logs=True,
    on_queue_update=on_queue_update,
)
print(result)

Input Schema

prompt

string

required

The prompt to fill the masked part of the image.

seed

integer

The same seed and the same prompt given to the same version of the model will output the same image every time.

sync_mode

boolean

default:"false"

If True, the media will be returned as a data URI and the output data won’t be available in the request history.

num_images

integer

default:"1"

The number of images to generate. Default value: 1Range: 1 to 4

output_format

OutputFormatEnum

default:"jpeg"

The format of the generated image. Default value: "jpeg"Possible values: jpeg, png

safety_tolerance

SafetyToleranceEnum

default:"2"

The safety tolerance level for the generated image. 1 being the most strict and 5 being the most permissive. Default value: "2"Possible values: 1, 2, 3, 4, 5, 6

enhance_prompt

boolean

default:"false"

Whether to enhance the prompt for better results.

image_url

string

required

The image URL to generate an image from. Needs to match the dimensions of the mask.

mask_url

string

required

The mask URL to inpaint the image. Needs to match the dimensions of the input image.

Output Schema

images

list<registry__image__fast_sdxl__models__Image>

required

The generated image files info.

timings

Timings

required

seed

integer

required

Seed of the generated Image. It will be the same value of the one passed in the input or the randomly generated that was used in case none was passed.

has_nsfw_concepts

list<boolean>

required

Whether the generated images contain NSFW concepts.

prompt

string

required

The prompt used for generating the image.

Input Example

{
  "prompt": "A knight in shining armour holding a greatshield with \"FAL\" on it",
  "sync_mode": false,
  "num_images": 1,
  "output_format": "jpeg",
  "safety_tolerance": "2",
  "enhance_prompt": false,
  "image_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/knight.jpeg",
  "mask_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/mask_knight.jpeg"
}

Output Example

{
  "images": [
    {
      "url": "",
      "content_type": "image/jpeg"
    }
  ],
  "prompt": ""
}

Endpoint: POST https://fal.run/fal-ai/flux-pro/v1/fill-finetuned Endpoint ID: fal-ai/flux-pro/v1/fill-finetuned

Try it in the Playground

Run this model interactively with your own prompts.

Quick Start

import fal_client

def on_queue_update(update):
    if isinstance(update, fal_client.InProgress):
        for log in update.logs:
           print(log["message"])

result = fal_client.subscribe(
    "fal-ai/flux-pro/v1/fill-finetuned",
    arguments={
        "prompt": "A knight in shining armour holding a greatshield with \"FAL\" on it",
        "image_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/knight.jpeg",
        "mask_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/mask_knight.jpeg",
        "finetune_id": ""
    },
    with_logs=True,
    on_queue_update=on_queue_update,
)
print(result)

Input Schema

prompt

string

required

The prompt to fill the masked part of the image.

seed

integer

The same seed and the same prompt given to the same version of the model will output the same image every time.

sync_mode

boolean

default:"false"

If True, the media will be returned as a data URI and the output data won’t be available in the request history.

num_images

integer

default:"1"

The number of images to generate. Default value: 1Range: 1 to 4

output_format

OutputFormatEnum

default:"jpeg"

The format of the generated image. Default value: "jpeg"Possible values: jpeg, png

safety_tolerance

SafetyToleranceEnum

default:"2"

The safety tolerance level for the generated image. 1 being the most strict and 5 being the most permissive. Default value: "2"Possible values: 1, 2, 3, 4, 5, 6

enhance_prompt

boolean

default:"false"

Whether to enhance the prompt for better results.

image_url

string

required

The image URL to generate an image from. Needs to match the dimensions of the mask.

mask_url

string

required

The mask URL to inpaint the image. Needs to match the dimensions of the input image.

finetune_id

string

required

References your specific model

finetune_strength

float

required

Controls finetune influence. Increase this value if your target concept isn’t showing up strongly enough. The optimal setting depends on your finetune and promptRange: 0 to 2

Output Schema

images

list<registry__image__fast_sdxl__models__Image>

required

The generated image files info.

timings

Timings

required

seed

integer

required

Seed of the generated Image. It will be the same value of the one passed in the input or the randomly generated that was used in case none was passed.

has_nsfw_concepts

list<boolean>

required

Whether the generated images contain NSFW concepts.

prompt

string

required

The prompt used for generating the image.

Input Example

{
  "prompt": "A knight in shining armour holding a greatshield with \"FAL\" on it",
  "sync_mode": false,
  "num_images": 1,
  "output_format": "jpeg",
  "safety_tolerance": "2",
  "enhance_prompt": false,
  "image_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/knight.jpeg",
  "mask_url": "https://storage.googleapis.com/falserverless/flux-lora/example-images/mask_knight.jpeg",
  "finetune_id": ""
}

Output Example

{
  "images": [
    {
      "url": "",
      "content_type": "image/jpeg"
    }
  ],
  "prompt": ""
}

Black Forest Labs’ FLUX.1 [pro] delivers professional-grade inpainting at $0.05 per megapixel. Trading broad creative generation for surgical precision, this endpoint excels at targeted modifications, replacing objects, filling gaps, or seamlessly blending new elements into existing compositions without regenerating entire images. Use Cases: Product photo editing | Architectural visualization refinement | E-commerce asset modification

Performance

FLUX.1 [pro] Fill operates at competitive inference speeds while maintaining the quality standards of the full FLUX.1 [pro] architecture, with cost scaling directly to output resolution rather than flat per-inference pricing.

Metric	Result	Context
Cost per Megapixel	$0.05	Billed by rounding up to nearest megapixel
Output Range	1-4 images	Single API call supports batch generation
Safety Tolerance	6 levels (1-6)	Configurable content moderation strictness
Output Formats	JPEG, PNG	Standard web-optimized formats
Related Endpoints	FLUX.1 Kontext [pro], FLUX.1 Kontext [max]	Context-aware variants for multi-reference workflows

Precision Inpainting Without Full Regeneration

FLUX.1 [pro] Fill operates through mask-based editing rather than full image regeneration, requiring three inputs: your base image, a mask defining the edit region, and a text prompt describing the desired fill content. This approach preserves everything outside the mask untouched while applying FLUX.1 [pro]‘s generation quality only where needed. What this means for you:

Surgical precision: Modify specific objects or regions without risking changes to surrounding elements, critical for product photography where brand consistency matters
Dual-input control: Mask URL and image URL parameters enable programmatic workflows where masks are generated separately (edge detection, object segmentation, manual selection tools)
Prompt enhancement option: Optional enhance_prompt parameter applies FLUX’s prompt optimization to your instructions, improving interpretation of complex editing requests
Batch efficiency: Generate up to 4 variations per API call with different seeds, useful for A/B testing edits or exploring alternative fills for the same masked region

Technical Specifications

Spec	Details
Architecture	FLUX.1 [pro]
Input Formats	Image URL (JPEG, PNG, WebP, GIF, AVIF) + Mask URL (matching dimensions)
Output Formats	JPEG, PNG
Resolution Handling	Variable (billed per megapixel, rounded up)
License	Commercial use permitted via fal partnership

API Documentation | Quickstart Guide | Enterprise Pricing

How It Stacks Up

**FLUX.1 Kontext [pro] (

0.055/MP)** – FLUX.1 [pro] Fill (

0.05/MP) focuses on single-mask inpainting with straightforward prompt-based fills at 9% lower cost. FLUX.1 Kontext [pro] adds multi-reference image conditioning via URL arrays, enabling style transfer and context-aware fills for workflows requiring visual examples beyond text prompts. **FLUX.1 Kontext [max] (

0.11/MP)** – FLUX.1 [pro] Fill delivers core inpainting functionality at 55% lower cost (

0.05 vs $0.11 per megapixel). FLUX.1 Kontext [max] trades cost efficiency for maximum quality output and enhanced multi-reference capabilities, ideal for high-end commercial work where generation fidelity justifies premium pricing.

FLUX1.1 [pro] ultra — Image Generation
FLUX1.1 [pro] — Image Generation
FLUX1.1 [pro] Redux — Image Generation
FLUX1.1 [pro] ultra Redux — Image Generation
FLUX1.1 [pro] ultra Fine-tuned — Image Generation
FLUX.1 [pro] Fill Fine-tuned — Image Generation
FLUX.1 [pro] Fill — Image Generation

Limitations

num_images range: 1 to 4
output_format restricted to: jpeg, png
safety_tolerance restricted to: 1, 2, 3, 4, 5, 6
finetune_strength range: 0 to 2

Video Generation

Image Generation

Audio

Vision

3D

Flux Pro V1 API

Try it in the Playground

Quick Start

Input Schema

Output Schema

Input Example

Output Example

Try it in the Playground

Quick Start

Input Schema

Output Schema

Input Example

Output Example

Performance

Precision Inpainting Without Full Regeneration

Technical Specifications

How It Stacks Up

Limitations

Video Generation

Image Generation

Audio

Vision

3D

Try it in the Playground

​Quick Start

​Input Schema

​Output Schema

​Input Example

​Output Example

Try it in the Playground

​Quick Start

​Input Schema

​Output Schema

​Input Example

​Output Example

​Performance

​Precision Inpainting Without Full Regeneration

​Technical Specifications

​How It Stacks Up

​Related

​Limitations

Quick Start

Input Schema

Output Schema

Input Example

Output Example

Quick Start

Input Schema

Output Schema

Input Example

Output Example

Performance

Precision Inpainting Without Full Regeneration

Technical Specifications

How It Stacks Up

Related

Limitations