fofr / aura-flow

Largest completely open sourced flow-based generation model that is capable of text-to-image generation

Cold

Public
7.5K runs
A100 (80GB)
GitHub
Paper
License

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

a stylized cartoon image, yellow background, a smiling turtle holding a pizza, the turtle is wearing a white and brown cap, large eyes and a friendly expression, the turtle is standing on its hind legs, the pizza is held in both of its front legs, small clouds and trees on either side of the turtle, the text "Turtle Pizza" is in large brown letters at the top, the text "No More Hungry" at the bottom in brown, the overall style is retro with halftone shadinga stylized cartoon image, yellow background, a smiling turtle holding a pizza, the turtle is wearing a white and brown cap, large eyes and a friendly expression, the turtle is standing on its hind legs, the pizza is held in both of its front legs, small clouds and trees on either side of the turtle, the text "Turtle Pizza" is in large brown letters at the top, the text "No More Hungry" at the bottom in brown, the overall style is retro with halftone shading

Default: ""

negative_prompt

string

Shift + Return to add a new line

Things you do not want to see in your image

Default: ""

width

integer

(minimum: 512, maximum: 2048)

The width of the image

Default: 1024

height

integer

(minimum: 512, maximum: 2048)

The height of the image

Default: 1024

number_of_images

integer

(minimum: 1, maximum: 10)

The number of images to generate

Default: 1

steps

integer

(minimum: 1, maximum: 100)

The number of steps to run the model for (more steps = better image but slower generation. Best results for this model are around 25 steps.)

Default: 25

cfg

number

(minimum: 0, maximum: 20)

The guidance scale tells the model how similar the output should be to the prompt.

Default: 3.5

sampler

string

An enumeration.

Default: "uni_pc"

scheduler

string

An enumeration.

Default: "normal"

shift

number

(minimum: 0, maximum: 10)

The timestep scheduling shift; shift values higher than 1.0 are better at managing noise in higher resolutions.

Default: 1.73

output_format

string

Format of the output images

Default: "webp"

output_quality

integer

(minimum: 0, maximum: 100)

Quality of the output images, from 0 to 100. 100 is best quality, 0 is lowest quality.

Default: 80

seed

integer

Set a seed for reproducibility. Random by default.

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

npx create-replicate --model=fofr/aura-flow

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 fofr/aura-flow using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "fofr/aura-flow:ae5ab66a7d1ca7ee44cf8c50265d3bafdef23734d03d66063d1c8fcf82f0c17b",
  {
    input: {
      cfg: 3.5,
      shift: 1.73,
      steps: 25,
      width: 1024,
      height: 1024,
      prompt: "a stylized cartoon image, yellow background, a smiling turtle holding a pizza, the turtle is wearing a white and brown cap, large eyes and a friendly expression, the turtle is standing on its hind legs, the pizza is held in both of its front legs, small clouds and trees on either side of the turtle, the text \"Turtle Pizza\" is in large brown letters at the top, the text \"No More Hungry\" at the bottom in brown, the overall style is retro with halftone shading",
      sampler: "uni_pc",
      scheduler: "normal",
      output_format: "webp",
      output_quality: 80,
      negative_prompt: "ugly, broken",
      number_of_images: 1
    }
  }
);

// 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 fofr/aura-flow using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fofr/aura-flow:ae5ab66a7d1ca7ee44cf8c50265d3bafdef23734d03d66063d1c8fcf82f0c17b",
    input={
        "cfg": 3.5,
        "shift": 1.73,
        "steps": 25,
        "width": 1024,
        "height": 1024,
        "prompt": "a stylized cartoon image, yellow background, a smiling turtle holding a pizza, the turtle is wearing a white and brown cap, large eyes and a friendly expression, the turtle is standing on its hind legs, the pizza is held in both of its front legs, small clouds and trees on either side of the turtle, the text \"Turtle Pizza\" is in large brown letters at the top, the text \"No More Hungry\" at the bottom in brown, the overall style is retro with halftone shading",
        "sampler": "uni_pc",
        "scheduler": "normal",
        "output_format": "webp",
        "output_quality": 80,
        "negative_prompt": "ugly, broken",
        "number_of_images": 1
    }
)
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 fofr/aura-flow 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": "ae5ab66a7d1ca7ee44cf8c50265d3bafdef23734d03d66063d1c8fcf82f0c17b",
    "input": {
      "cfg": 3.5,
      "shift": 1.73,
      "steps": 25,
      "width": 1024,
      "height": 1024,
      "prompt": "a stylized cartoon image, yellow background, a smiling turtle holding a pizza, the turtle is wearing a white and brown cap, large eyes and a friendly expression, the turtle is standing on its hind legs, the pizza is held in both of its front legs, small clouds and trees on either side of the turtle, the text \\"Turtle Pizza\\" is in large brown letters at the top, the text \\"No More Hungry\\" at the bottom in brown, the overall style is retro with halftone shading",
      "sampler": "uni_pc",
      "scheduler": "normal",
      "output_format": "webp",
      "output_quality": 80,
      "negative_prompt": "ugly, broken",
      "number_of_images": 1
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2024-07-12T09:31:35.354499Z",
  "created_at": "2024-07-12T09:31:10.252000Z",
  "data_removed": false,
  "error": null,
  "id": "tf0mz49w5hrgm0cgms4byp74gr",
  "input": {
    "cfg": 3.5,
    "shift": 1.73,
    "steps": 25,
    "width": 1024,
    "height": 1024,
    "prompt": "a stylized cartoon image, yellow background, a smiling turtle holding a pizza, the turtle is wearing a white and brown cap, large eyes and a friendly expression, the turtle is standing on its hind legs, the pizza is held in both of its front legs, small clouds and trees on either side of the turtle, the text \"Turtle Pizza\" is in large brown letters at the top, the text \"No More Hungry\" at the bottom in brown, the overall style is retro with halftone shading",
    "sampler": "uni_pc",
    "scheduler": "normal",
    "output_format": "webp",
    "output_quality": 80,
    "negative_prompt": "ugly, broken",
    "number_of_images": 1
  },
  "logs": "Random seed set to: 4198727867\nRunning workflow\ngot prompt\nExecuting node 3, title: KSampler, class type: KSampler\n  0%|          | 0/25 [00:00<?, ?it/s]\n  4%|▍         | 1/25 [00:00<00:04,  5.30it/s]\n  8%|▊         | 2/25 [00:01<00:14,  1.56it/s]\n 12%|█▏        | 3/25 [00:02<00:17,  1.27it/s]\n 16%|█▌        | 4/25 [00:03<00:17,  1.17it/s]\n 20%|██        | 5/25 [00:04<00:17,  1.12it/s]\n 24%|██▍       | 6/25 [00:04<00:17,  1.09it/s]\n 28%|██▊       | 7/25 [00:05<00:16,  1.07it/s]\n 32%|███▏      | 8/25 [00:06<00:15,  1.06it/s]\n 36%|███▌      | 9/25 [00:07<00:15,  1.06it/s]\n 40%|████      | 10/25 [00:08<00:14,  1.05it/s]\n 44%|████▍     | 11/25 [00:09<00:13,  1.05it/s]\n 48%|████▊     | 12/25 [00:10<00:12,  1.05it/s]\n 52%|█████▏    | 13/25 [00:11<00:11,  1.05it/s]\n 56%|█████▌    | 14/25 [00:12<00:10,  1.04it/s]\n 60%|██████    | 15/25 [00:13<00:09,  1.04it/s]\n 64%|██████▍   | 16/25 [00:14<00:08,  1.04it/s]\n 68%|██████▊   | 17/25 [00:15<00:07,  1.04it/s]\n 72%|███████▏  | 18/25 [00:16<00:06,  1.04it/s]\n 76%|███████▌  | 19/25 [00:17<00:05,  1.04it/s]\n 80%|████████  | 20/25 [00:18<00:04,  1.04it/s]\n 84%|████████▍ | 21/25 [00:19<00:03,  1.04it/s]\n 88%|████████▊ | 22/25 [00:20<00:02,  1.04it/s]\n 92%|█████████▏| 23/25 [00:21<00:01,  1.04it/s]\n 96%|█████████▌| 24/25 [00:22<00:00,  1.04it/s]\n100%|██████████| 25/25 [00:24<00:00,  1.19s/it]\n100%|██████████| 25/25 [00:24<00:00,  1.04it/s]\nExecuting node 8, title: VAE Decode, class type: VAEDecode\nExecuting node 9, title: Save Image, class type: SaveImage\nPrompt executed in 24.38 seconds\noutputs:  {'9': {'images': [{'filename': 'ComfyUI_00001_.png', 'subfolder': '', 'type': 'output'}]}}\n====================================\nComfyUI_00001_.png",
  "metrics": {
    "predict_time": 25.089343742,
    "total_time": 25.102499
  },
  "output": [
    "https://replicate.delivery/pbxt/FB6BVEL8rUaaKpR6oCfd2wepBKgXyb3ay93qo7WuCequPVPmA/ComfyUI_00001_.webp"
  ],
  "started_at": "2024-07-12T09:31:10.265155Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/tf0mz49w5hrgm0cgms4byp74gr",
    "cancel": "https://api.replicate.com/v1/predictions/tf0mz49w5hrgm0cgms4byp74gr/cancel"
  },
  "version": "2f38ca7d2e0e84f05bef7980cf8c7c1c36ba2b9416eee3d57e73b67c338eaff0"
}

Generated in

25.1 seconds

Tweak it ShareReport

Random seed set to: 4198727867
Running workflow
got prompt
Executing node 3, title: KSampler, class type: KSampler
  0%|          | 0/25 [00:00<?, ?it/s]
  4%|▍         | 1/25 [00:00<00:04,  5.30it/s]
  8%|▊         | 2/25 [00:01<00:14,  1.56it/s]
 12%|█▏        | 3/25 [00:02<00:17,  1.27it/s]
 16%|█▌        | 4/25 [00:03<00:17,  1.17it/s]
 20%|██        | 5/25 [00:04<00:17,  1.12it/s]
 24%|██▍       | 6/25 [00:04<00:17,  1.09it/s]
 28%|██▊       | 7/25 [00:05<00:16,  1.07it/s]
 32%|███▏      | 8/25 [00:06<00:15,  1.06it/s]
 36%|███▌      | 9/25 [00:07<00:15,  1.06it/s]
 40%|████      | 10/25 [00:08<00:14,  1.05it/s]
 44%|████▍     | 11/25 [00:09<00:13,  1.05it/s]
 48%|████▊     | 12/25 [00:10<00:12,  1.05it/s]
 52%|█████▏    | 13/25 [00:11<00:11,  1.05it/s]
 56%|█████▌    | 14/25 [00:12<00:10,  1.04it/s]
 60%|██████    | 15/25 [00:13<00:09,  1.04it/s]
 64%|██████▍   | 16/25 [00:14<00:08,  1.04it/s]
 68%|██████▊   | 17/25 [00:15<00:07,  1.04it/s]
 72%|███████▏  | 18/25 [00:16<00:06,  1.04it/s]
 76%|███████▌  | 19/25 [00:17<00:05,  1.04it/s]
 80%|████████  | 20/25 [00:18<00:04,  1.04it/s]
 84%|████████▍ | 21/25 [00:19<00:03,  1.04it/s]
 88%|████████▊ | 22/25 [00:20<00:02,  1.04it/s]
 92%|█████████▏| 23/25 [00:21<00:01,  1.04it/s]
 96%|█████████▌| 24/25 [00:22<00:00,  1.04it/s]
100%|██████████| 25/25 [00:24<00:00,  1.19s/it]
100%|██████████| 25/25 [00:24<00:00,  1.04it/s]
Executing node 8, title: VAE Decode, class type: VAEDecode
Executing node 9, title: Save Image, class type: SaveImage
Prompt executed in 24.38 seconds
outputs:  {'9': {'images': [{'filename': 'ComfyUI_00001_.png', 'subfolder': '', 'type': 'output'}]}}
====================================
ComfyUI_00001_.png

This output was created using a different version of the model, fofr/aura-flow:2f38ca7d.

Examples

View more examples

Run time and cost

This model costs approximately $0.054 to run on Replicate, or 18 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 39 seconds. The predict time for this model varies significantly based on the inputs.

Readme

AuraFlow

The largest yet completely open sourced flow-based generation model that is capable of text-to-image generation. Created by @cloneofsimo and @fal.

Blog post:

https://blog.fal.ai/auraflow/

Weights:

https://huggingface.co/fal/AuraFlow