POST
/
texture-mesh
/
project
/
create
Header
Body
curl -X 'POST' \
  'https://api.mirageml.com/texture-mesh/project/create' \
  -H 'Authorization: <AUTHORIZATION>' \
  -H 'Cache-Control: no-cache' \
  -H 'content-type: multipart/form-data; boundary=----WebKitFormBoundary7MA4YWxkTrZu0gW' \
  -H 'x-api-key: <API_KEY>' \
  -F prompt=<PROMPT> \
  -F 'files[]=<FILES>' \

{
  "statusCode": 200,
  "project_id": "66bb509b-7888-427a-b9fc-1a915a46fba0"
}

​
Description

Create a mesh with prompt and mesh algorithm

​
Form Data

prompt
string
required

The input prompt.

Enum (get3d): car, motorbike, chair, table, rocket, airplane, train

mesh-algorithm
string
default: "point-e"

Enum: point-e, get3d, dreamfusion

The underlying algorithm to use for generating the mesh.

guidance
string
default: "base1B"

Enum (point-e): base40M, base300M, base1B

Number of parameters used to create model.

Enum (dreamfusion): clip, stable-diffusion

CLIP and Stable Diffusion are text-to-image embedding models used.

train-iterations
number
default: "3"

Point-E: How much the image impacts the final output

Dreamfusion: 100 iterations are recommended for highest quality results, fewer iterations will finish faster

stable-diffusion-checkpoint
string
default: "stabilityai/stable-diffusion-2-1-base"

Change the stable diffusion checkpoint. You can use any public Hugging Face Checkpoint.

polygon
string
default: "tris"

Enum: tris, quads

The type of polygons to use when generating the mesh

negative-prompt
string

Use Negative Prompt to suppress output

links
string

JSON object of links to images to download. Example:

{"0": "https://sabe.io/images/saturn.png" }
files[]
files
Get ProjectUpload Project