Text to 3D

Use model_generate_fromprompt when you want to generate a 3D model from a text description.

When To Use

idea exploration from prompt-only inputs
rapid concept generation
no source image or source model available

Expected outcome: one or more GLB outputs retrievable by asset_id after completion.

Inputs

required input data type: text prompt string
recommended: concrete subject, style, and constraints in one prompt

Minimal input object:

{
  "data": "A modern office chair with armrests"
}

Parameters

Name	Type	Default	Allowed values	Practical guidance
`quality`	string	`high`	`low`, `high`	Use `high` for better geometry; use `low` for faster experiments.
`seed`	integer	`-1`	integer	Use fixed value for reproducibility; use `-1` for random runs.
`texture`	boolean	`true`	`true`, `false`	Set `true` to include textures.

Minimal Runnable Payload

payload = {
    "template": "model_generate_fromprompt",
    "parameters": {
        "quality": "high",
        "texture": True,
        "seed": -1
    },
    "inputs": [
        {
            "data": "A modern office chair with armrests"
        }
    ]
}

Polling And Output Retrieval

Create flow with POST /flows.
Poll GET /flows/{flow_id} until terminal state.
If completed, list outputs via GET /flows/{flow_id}/outputs.
Fetch asset with GET /flows/{flow_id}/outputs/{asset_id}?include_data=true.

For reusable polling/download code, use Template Helpers.

Common Failure Modes

empty or vague prompt resulting in poor output quality
invalid template field or malformed payload
output fetch before completion

Recovery:

strengthen prompt specificity
validate payload shape
retry output retrieval only after completed

See Troubleshooting for HTTP and flow-state error patterns.