jhorovitz/flux-dev-redux-control – Run with an API on Replicate

jhorovitz / flux-dev-redux-control

This models allow changing the strength of the Redux image prompt, which allows the text prompt to have a stronger effect. It is particularly useful at taking content from the provided image and applying style or editing changes from the prompt.

Cold

Public
1.5K runs
H100
GitHub
License

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

Text prompt

redux

file

Preview

Redux image.

redux_strength

number

(minimum: 0, maximum: 1)

Strength of the Redux image, values between 0.01 and 0.1 tend to have good results.

Default: 0.05

control_image

file

Control image (optional).

width

integer

Width, in pixels

Default: 1024

height

integer

Height, in pixels

Default: 1024

num_outputs

integer

(minimum: 1, maximum: 4)

Number of images to output.

Default: 1

num_inference_steps

integer

(minimum: 1, maximum: 50)

Number of inference steps

Default: 28

guidance_scale

number

(minimum: 0, maximum: 10)

Guidance scale for the diffusion process

Default: 3.5

seed

integer

Random seed. Set for reproducible generation

output_format

string

Format of the output images

Default: "webp"

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: 80

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

npx create-replicate --model=jhorovitz/flux-dev-redux-control

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 jhorovitz/flux-dev-redux-control using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "jhorovitz/flux-dev-redux-control:2d77f82dce2466dee3646296420ef620bf8ab28580ef4cd7c60d22d8900e5337",
  {
    input: {
      seed: 42,
      redux: "https://replicate.delivery/pbxt/MHoZIY0XUcZ2G34UPSRRXGOxOarxnMDc96u7UCrGdUsqoC1j/car.webp",
      width: 1024,
      height: 1024,
      prompt: "A red car driving underwater.",
      num_outputs: 1,
      output_format: "webp",
      guidance_scale: 3.5,
      output_quality: 80,
      redux_strength: 0.01,
      num_inference_steps: 28
    }
  }
);

// To access the file URL:
console.log(output[0].url()); //=> "http://example.com"

// To write the file to disk:
fs.writeFile("my-image.png", output[0]);

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 jhorovitz/flux-dev-redux-control using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "jhorovitz/flux-dev-redux-control:2d77f82dce2466dee3646296420ef620bf8ab28580ef4cd7c60d22d8900e5337",
    input={
        "seed": 42,
        "redux": "https://replicate.delivery/pbxt/MHoZIY0XUcZ2G34UPSRRXGOxOarxnMDc96u7UCrGdUsqoC1j/car.webp",
        "width": 1024,
        "height": 1024,
        "prompt": "A red car driving underwater.",
        "num_outputs": 1,
        "output_format": "webp",
        "guidance_scale": 3.5,
        "output_quality": 80,
        "redux_strength": 0.01,
        "num_inference_steps": 28
    }
)
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 jhorovitz/flux-dev-redux-control 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": "jhorovitz/flux-dev-redux-control:2d77f82dce2466dee3646296420ef620bf8ab28580ef4cd7c60d22d8900e5337",
    "input": {
      "seed": 42,
      "redux": "https://replicate.delivery/pbxt/MHoZIY0XUcZ2G34UPSRRXGOxOarxnMDc96u7UCrGdUsqoC1j/car.webp",
      "width": 1024,
      "height": 1024,
      "prompt": "A red car driving underwater.",
      "num_outputs": 1,
      "output_format": "webp",
      "guidance_scale": 3.5,
      "output_quality": 80,
      "redux_strength": 0.01,
      "num_inference_steps": 28
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2025-01-08T15:02:11.115402Z",
  "created_at": "2025-01-08T15:01:32.978000Z",
  "data_removed": false,
  "error": null,
  "id": "5qc07hk4p9rme0cm8sqt3tvr7r",
  "input": {
    "seed": 42,
    "redux": "https://replicate.delivery/pbxt/MHoZIY0XUcZ2G34UPSRRXGOxOarxnMDc96u7UCrGdUsqoC1j/car.webp",
    "width": 1024,
    "height": 1024,
    "prompt": "A red car driving underwater.",
    "num_outputs": 1,
    "output_format": "webp",
    "guidance_scale": 3.5,
    "output_quality": 80,
    "redux_strength": 0.01,
    "num_inference_steps": 28
  },
  "logs": "Using seed: 42\n  0%|          | 0/28 [00:00<?, ?it/s]\n  4%|▎         | 1/28 [00:00<00:24,  1.11it/s]\n  7%|▋         | 2/28 [00:01<00:19,  1.32it/s]\n 11%|█         | 3/28 [00:02<00:20,  1.21it/s]\n 14%|█▍        | 4/28 [00:03<00:20,  1.17it/s]\n 18%|█▊        | 5/28 [00:04<00:20,  1.15it/s]\n 21%|██▏       | 6/28 [00:05<00:19,  1.13it/s]\n 25%|██▌       | 7/28 [00:06<00:18,  1.12it/s]\n 29%|██▊       | 8/28 [00:06<00:17,  1.12it/s]\n 32%|███▏      | 9/28 [00:07<00:17,  1.12it/s]\n 36%|███▌      | 10/28 [00:08<00:16,  1.11it/s]\n 39%|███▉      | 11/28 [00:09<00:15,  1.11it/s]\n 43%|████▎     | 12/28 [00:10<00:14,  1.11it/s]\n 46%|████▋     | 13/28 [00:11<00:13,  1.11it/s]\n 50%|█████     | 14/28 [00:12<00:12,  1.11it/s]\n 54%|█████▎    | 15/28 [00:13<00:11,  1.11it/s]\n 57%|█████▋    | 16/28 [00:14<00:10,  1.11it/s]\n 61%|██████    | 17/28 [00:15<00:09,  1.11it/s]\n 64%|██████▍   | 18/28 [00:16<00:09,  1.11it/s]\n 68%|██████▊   | 19/28 [00:16<00:08,  1.11it/s]\n 71%|███████▏  | 20/28 [00:17<00:07,  1.11it/s]\n 75%|███████▌  | 21/28 [00:18<00:06,  1.11it/s]\n 79%|███████▊  | 22/28 [00:19<00:05,  1.11it/s]\n 82%|████████▏ | 23/28 [00:20<00:04,  1.11it/s]\n 86%|████████▌ | 24/28 [00:21<00:03,  1.11it/s]\n 89%|████████▉ | 25/28 [00:22<00:02,  1.11it/s]\n 93%|█████████▎| 26/28 [00:23<00:01,  1.11it/s]\n 96%|█████████▋| 27/28 [00:24<00:00,  1.11it/s]\n100%|██████████| 28/28 [00:25<00:00,  1.11it/s]\n100%|██████████| 28/28 [00:25<00:00,  1.12it/s]",
  "metrics": {
    "predict_time": 25.64457902,
    "total_time": 38.137402
  },
  "output": [
    "https://replicate.delivery/xezq/XSlusWLDzMbeTaIVZb4A195GwHECZut72oO0AhMp8e6zVEDUA/out-0.webp"
  ],
  "started_at": "2025-01-08T15:01:45.470823Z",
  "status": "succeeded",
  "urls": {
    "stream": "https://stream.replicate.com/v1/files/bcwr-6d4keee75a74hmj6tesj6schc7kdjq6w3dtjihhvqo3fvzh4s2fq",
    "get": "https://api.replicate.com/v1/predictions/5qc07hk4p9rme0cm8sqt3tvr7r",
    "cancel": "https://api.replicate.com/v1/predictions/5qc07hk4p9rme0cm8sqt3tvr7r/cancel"
  },
  "version": "c7e5394e7fc2b92760d37d607e07af5368ef4a61e59c40049aaa71e76c714430"
}

Generated in

25.6 seconds

Tweak it Share Report

This output was created using a different version of the model, jhorovitz/flux-dev-redux-control:c7e5394e.

Examples

View more examples

Run time and cost

This model costs approximately $0.17 to run on Replicate, or 5 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 H100 GPU hardware. Predictions typically complete within 110 seconds. The predict time for this model varies significantly based on the inputs.

Readme

This model doesn't have a readme.