3DGen AI

Generate 3d models with image (Image-to-3D). See your generated 3D model in seconds, and export as 3d file.

Note: To generate 3d models with only prompt (Text-to-3D), we suggest you to use Fluw AI and get the image from this endpoint to use with 3DGen AI endpoint.

POST /v2/generate

Parameters

Name
Type
Description

model

string

Required

Value must be 3dgen for 3DGen AI. See other generators. Browse other Generators.

image

string

Optional

URL of the input image to get 3d model of image file. Encode your image url if url have special characters.

extra_image_1

string

Optional

URL of the extra input image 1 to get 3d model. Side photos & angles

extra_image_2

string

Optional

URL of the extra input image 2 to get 3d model. Side photos & angles

extra_image_3

string

Optional

URL of the extra input image 3 to get 3d model. Side photos & angles

seed

integer

Optional

For reproducible results

luwmodel

string

Optional

If you want to use Luw 3 Symphony model, use symphony-3 value. If you want to continue using Luw 2.5 Aria model don't add luwmodel parameter to your requests.

pid

integer

Optional

Persona ID to use if you want to use Persona Slots for images.

processing_url

string

Optional

Get process results with same endpoint, alternative to /results endpoint usage.

webhook

string

Optional

Get process results with real-time webhooks, alternative to /results endpoint usage.

If you want to use additional images to generate 3D model better, also add extra images: Visual training slots from 1 to 3 or use extra_image_1 extra_image_2 and extra_image_3 parameters

Curl Example

curl -X POST https://api.luw.ai/v2/generate \
  -H "Authorization: Bearer LUW_API_TOKEN" \
  -d "model=3dgen" \
  -d "image=https://example.com/sofa.jpg"

Response Example

{
  "status": true,
  "progress": {
    "started_at": "2024-01-04T12:00:00Z",
    "state": "processing",
    "percent": 45
  },
  "processing": true,
  "processing_url": "pred_123..."
}

✨ Get results using the /results endpoint.

Alternative legacy method: Add &processing_url={url} to your request.

Last updated