Generate 3D Model
Generate a 3D model from a text prompt or reference image(s). Returns GLB by default. nvidia/asset-harvester returns PLY.
Available models: microsoft/trellis-2, microsoft/trellis-2:fal, nvidia/asset-harvester, hyper3d/rodin-2.5. microsoft/trellis-2 is the default.
Pass reference image URL(s) via the image parameter for image-to-3D models (microsoft/trellis-2, nvidia/asset-harvester). Separate multiple URLs with | or ,. hyper3d/rodin-2.5 accepts both images and a text prompt.
Browse all available models and their input requirements at /3d/models.
Request
- Base URL: https://gen.pollinations.ai
- URL: https://gen.pollinations.ai/3d/{prompt}
- Auth: HTTP bearer
Path parameters
Text description of the 3D model to generate (required for text-to-3D models such as Hyper3D Rodin; ignored by image-only models such as Trellis 2)
Query parameters
Model to use. See /3d/models for the full list and per-model input requirements.
Output detail for microsoft/trellis-2. Defaults to low.
Reference image URL(s) for image-to-3D generation. Separate multiple URLs with | or ,. Required for image-only models (e.g. trellis, triposr, sf3d).
Seed for varied generations. Passed through to models that support it (hyper3d/rodin-2.5); otherwise only affects the media-cache key, so a new seed forces a fresh generation for the same prompt/image.
Safety features: comma-separated list of privacy, secrets, sexual, violence, shield, true, nsfw. true enables privacy,secrets; nsfw enables sexual,violence. Also accepted in the Pollinations-Safe header. Defaults to off; false and 0 are accepted as off.
Response
Success - Returns the generated 3D model
Changes
No recorded changes to this operation across all 1 revision of this API.