Recraft V3

Recraft V3 is an image generation model released in October 2024 — a state-of-the-art model trained from scratch that introduced major advances in photorealism and text rendering. It ranked first on the Hugging Face Text-to-Image leaderboard for five consecutive months, maintaining a clear lead over models from Midjourney, OpenAI, and others. In the API, V3 remains the most widely supported model line: inpainting, outpainting, and the background operations run on V3 models only, and it is the only line that can place text at exact positions in an image.

Models

Model	`model` value	Output	Price per image
Recraft V3	`recraftv3`	Raster, 1K	$0.04
Recraft V3 Vector	`recraftv3_vector`	Vector, SVG	$0.08

How to call it

Pass the model to /v1/images/generations:

response = client.images.generate(
    prompt='two race cars on a track',
    model='recraftv3',
)
print(response.data[0].url)

curl -X POST https://external.api.recraft.ai/v1/images/generations \
  -H "Authorization: Bearer $RECRAFT_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "two race cars on a track",
    "model": "recraftv3"
  }'

Supported operations

V3 models are accepted by every operation below:

Operation	Endpoint	Price (raster / vector)
Generate image	`/v1/images/generations`	$0.04 /$ 0.08
Create style	`/v1/styles`	$0.005 per request
Image to image	`/v1/images/imageToImage`	$0.04 /$ 0.08
Image inpainting	`/v1/images/inpaint`	$0.04 /$ 0.08
Image outpainting	`/v1/images/outpaint`	$0.04 /$ 0.08
Replace background	`/v1/images/replaceBackground`	$0.04 /$ 0.08
Generate background	`/v1/images/generateBackground`	$0.04 /$ 0.08

Inpainting, outpainting, and the background operations are V3-exclusive; generate image and create style accept every model line, and image to image also accepts V4 and V4.1 models.

Additional features

Feature	Support
Vector generation	Editable SVG output: `recraftv3_vector`
Styles	A rich curated library, custom styles, `style_id`, and attached style references. Default style: `Recraft V3 Raw` (raster), `Vector art` (vector)
Text layout	The only models that place text at specific positions in an image
`negative_prompt`	A text description of undesired elements on an image
Controls	Full set, including the V3-only `artistic_level` and `no_text`

Endpoints

Full parameter reference for generation and every editing operation.

Pricing

Complete API unit charges across every model and operation.

Styles

The full curated style list for V3 and V3 Vector.

The model’s background and strengths in detail.

​Models

​How to call it

​Supported operations

​Additional features

​Related