fermatresearch/flux-controlnet-inpaint – Run with an API on Replicate

Input

prompt

*string

Shift + Return to add a new line

Professional photography of a fox sitting on a benchProfessional photography of a fox sitting on a bench

Prompt for generated image

conditioning_scale

number

(minimum: 0, maximum: 1)

ControlNet strength, depth works best at 0.2, canny works best at 0.4. Recommended range is 0.3-0.8

Default: 0.5

image

file

Preview

The image to restyle

control_image

file

Preview

The image to control the generation

mask

file

Preview

The area to inpaint

strength

number

(minimum: 0, maximum: 1)

Img2Img strength

Default: 0.8

guidance_scale

number

(minimum: 0, maximum: 30)

Guidance scale

Default: 3.5

enable_hyper_flux_8_step

boolean

Whether to use Hyper-FLUX.1-dev-8steps or not. If False, make sure to increase your number of inference steps

Default: true

num_inference_steps

integer

(minimum: 1, maximum: 28)

Number of inference steps

Default: 8

seed

integer

Random seed. Set for reproducible generation

output_format

string

Format of the output images

Default: "jpg"

output_quality

integer

(minimum: 0, maximum: 100)

Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs

Default: 100

lora_weights

string

Shift + Return to add a new line

Huggingface path, or URL to the LoRA weights. Ex: alvdansen/frosting_lane_flux

lora_scale

number

(minimum: 0, maximum: 1)

Scale for the LoRA weights

Default: 0.8

Run this model in Node.js with one line of code:

npx create-replicate --model=fermatresearch/flux-controlnet-inpaint

or set up a project from scratch

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import and set up the client:

import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

Run fermatresearch/flux-controlnet-inpaint using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "fermatresearch/flux-controlnet-inpaint:46ae77d1d148dca1bd61c1215e99ee692c4cf01289a2bba681b14a58c04bda8f",
  {
    input: {
      mask: "https://replicate.delivery/pbxt/LdzOC0eYgmiQxsZntL8aUwyQlLjtXQuhmMrXrOUbpMESw11v/output%20%282%29.png",
      image: "https://replicate.delivery/pbxt/LdzOC2USRj4omEK2igawxoYUPG6hJ2RfRZS2TZXiqSEkybYs/output%20%283%29.png",
      prompt: "Professional photography of a fox sitting on a bench",
      strength: 0.8,
      lora_scale: 0.8,
      control_image: "https://replicate.delivery/pbxt/LdzOBqGI9FyIIPZOyOBYToYndRLXelYBSqOP1XJlE5Tq5up9/output%20%283%29.png",
      output_format: "jpg",
      guidance_scale: 3.5,
      output_quality: 100,
      conditioning_scale: 0.5,
      num_inference_steps: 8,
      enable_hyper_flux_8_step: true
    }
  }
);
console.log(output);

To learn more, take a look at the guide on getting started with Node.js.

Install Replicate’s Python client library:

pip install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import the client:

import replicate

Run fermatresearch/flux-controlnet-inpaint using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fermatresearch/flux-controlnet-inpaint:46ae77d1d148dca1bd61c1215e99ee692c4cf01289a2bba681b14a58c04bda8f",
    input={
        "mask": "https://replicate.delivery/pbxt/LdzOC0eYgmiQxsZntL8aUwyQlLjtXQuhmMrXrOUbpMESw11v/output%20%282%29.png",
        "image": "https://replicate.delivery/pbxt/LdzOC2USRj4omEK2igawxoYUPG6hJ2RfRZS2TZXiqSEkybYs/output%20%283%29.png",
        "prompt": "Professional photography of a fox sitting on a bench",
        "strength": 0.8,
        "lora_scale": 0.8,
        "control_image": "https://replicate.delivery/pbxt/LdzOBqGI9FyIIPZOyOBYToYndRLXelYBSqOP1XJlE5Tq5up9/output%20%283%29.png",
        "output_format": "jpg",
        "guidance_scale": 3.5,
        "output_quality": 100,
        "conditioning_scale": 0.5,
        "num_inference_steps": 8,
        "enable_hyper_flux_8_step": True
    }
)
print(output)

To learn more, take a look at the guide on getting started with Python.

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Run fermatresearch/flux-controlnet-inpaint using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

curl -s -X POST \
  -H "Authorization: Bearer $REPLICATE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Prefer: wait" \
  -d $'{
    "version": "46ae77d1d148dca1bd61c1215e99ee692c4cf01289a2bba681b14a58c04bda8f",
    "input": {
      "mask": "https://replicate.delivery/pbxt/LdzOC0eYgmiQxsZntL8aUwyQlLjtXQuhmMrXrOUbpMESw11v/output%20%282%29.png",
      "image": "https://replicate.delivery/pbxt/LdzOC2USRj4omEK2igawxoYUPG6hJ2RfRZS2TZXiqSEkybYs/output%20%283%29.png",
      "prompt": "Professional photography of a fox sitting on a bench",
      "strength": 0.8,
      "lora_scale": 0.8,
      "control_image": "https://replicate.delivery/pbxt/LdzOBqGI9FyIIPZOyOBYToYndRLXelYBSqOP1XJlE5Tq5up9/output%20%283%29.png",
      "output_format": "jpg",
      "guidance_scale": 3.5,
      "output_quality": 100,
      "conditioning_scale": 0.5,
      "num_inference_steps": 8,
      "enable_hyper_flux_8_step": true
    }
  }' \
  https://api.replicate.com/v1/predictions

To learn more, take a look at Replicate’s HTTP API reference docs.

You can run this model locally using Cog. First, install Cog:

brew install cog

If you don’t have Homebrew, there are other installation options available.

Run this to download the model and run it in your local environment:

cog predict r8.im/fermatresearch/flux-controlnet-inpaint@sha256:46ae77d1d148dca1bd61c1215e99ee692c4cf01289a2bba681b14a58c04bda8f \
  -i 'mask="https://replicate.delivery/pbxt/LdzOC0eYgmiQxsZntL8aUwyQlLjtXQuhmMrXrOUbpMESw11v/output%20%282%29.png"' \
  -i 'image="https://replicate.delivery/pbxt/LdzOC2USRj4omEK2igawxoYUPG6hJ2RfRZS2TZXiqSEkybYs/output%20%283%29.png"' \
  -i 'prompt="Professional photography of a fox sitting on a bench"' \
  -i 'strength=0.8' \
  -i 'lora_scale=0.8' \
  -i 'control_image="https://replicate.delivery/pbxt/LdzOBqGI9FyIIPZOyOBYToYndRLXelYBSqOP1XJlE5Tq5up9/output%20%283%29.png"' \
  -i 'output_format="jpg"' \
  -i 'guidance_scale=3.5' \
  -i 'output_quality=100' \
  -i 'conditioning_scale=0.5' \
  -i 'num_inference_steps=8' \
  -i 'enable_hyper_flux_8_step=true'

To learn more, take a look at the Cog documentation.

Run this to download the model and run it in your local environment:

docker run -d -p 5000:5000 --gpus=all r8.im/fermatresearch/flux-controlnet-inpaint@sha256:46ae77d1d148dca1bd61c1215e99ee692c4cf01289a2bba681b14a58c04bda8f
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "mask": "https://replicate.delivery/pbxt/LdzOC0eYgmiQxsZntL8aUwyQlLjtXQuhmMrXrOUbpMESw11v/output%20%282%29.png",
      "image": "https://replicate.delivery/pbxt/LdzOC2USRj4omEK2igawxoYUPG6hJ2RfRZS2TZXiqSEkybYs/output%20%283%29.png",
      "prompt": "Professional photography of a fox sitting on a bench",
      "strength": 0.8,
      "lora_scale": 0.8,
      "control_image": "https://replicate.delivery/pbxt/LdzOBqGI9FyIIPZOyOBYToYndRLXelYBSqOP1XJlE5Tq5up9/output%20%283%29.png",
      "output_format": "jpg",
      "guidance_scale": 3.5,
      "output_quality": 100,
      "conditioning_scale": 0.5,
      "num_inference_steps": 8,
      "enable_hyper_flux_8_step": true
    }
  }' \
  http://localhost:5000/predictions

To learn more, take a look at the Cog documentation.

Output

Generated in

10.9 seconds

Tweak it ShareReport

Run time and cost

This model costs approximately $0.036 to run on Replicate, or 27 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on Nvidia A100 (80GB) GPU hardware. Predictions typically complete within 26 seconds. The predict time for this model varies significantly based on the inputs.