Image-to-3D mesh generation via Stability AI's Stable Point Aware 3D model, pay-per-call over MPP.
What it does
This endpoint provides access to Stability AI's Stable Point Aware 3D (SPA3D) model through the Locus MPP (Micropayment Protocol) gateway. Given a source image of an object, it generates a textured 3D mesh. The model supports configurable texture resolution (512, 1024, or 2048), foreground ratio scaling, guidance scale, and deterministic output via seed control. Payment is settled per call via the Tempo method on pathUSD.
The endpoint is priced at 46,000 base units of pathUSD (6 decimals), which equals $0.046 per generation. It is part of a broader Stability AI service suite available through the same Locus MPP gateway, which includes image generation (Ultra, Core, SD3), editing (erase, inpaint, outpaint, search-and-replace), 3D (Stable Fast 3D), upscaling, style transfer, and audio generation endpoints. The specific path for this endpoint is `/stability-ai/stable-point-aware-3d` and it accepts POST requests with a JSON body.
Note that the probe did not receive a 402 challenge on HEAD/GET — this is expected since the endpoint only responds to POST. The OpenAPI spec from the gateway confirms the endpoint exists with full schema and pricing metadata. The only required field is `image` (the source image of the object to reconstruct in 3D). Optional parameters include `texture_resolution`, `foreground_ratio`, `guidance_scale`, and `seed`. Some operations in this suite are asynchronous; a free `/stability-ai/result` endpoint is available to poll for completed generations.
Capabilities
Use cases
- —Generating 3D assets from product photos for e-commerce catalogs
- —Creating 3D models from concept art for game development pipelines
- —Rapid prototyping of 3D objects from reference images
- —Building 3D previews of real-world objects for AR/VR applications
- —Automating 3D asset creation in content production workflows
Fit
Best for
- —Single-image 3D mesh generation with texture
- —Agents needing pay-per-call 3D generation without API key management
- —Rapid 3D asset prototyping from 2D references
- —Programmatic 3D content pipelines settled via crypto micropayments
Not for
- —Multi-view or video-based 3D reconstruction
- —High-polygon production-ready 3D models requiring manual refinement
- —Real-time 3D generation with sub-second latency requirements
Quick start
curl -X POST https://stability-ai.mpp.paywithlocus.com/stability-ai/stable-point-aware-3d \
-H "Content-Type: application/json" \
-H "Authorization: <MPP_TEMPO_TOKEN>" \
-d '{"image": "<base64-encoded-image>"}'Example
Request
{
"seed": 42,
"image": "<base64-encoded-image-of-object>",
"guidance_scale": 3,
"foreground_ratio": 1.3,
"texture_resolution": 1024
}Endpoint
Quality
Full OpenAPI schema with request parameters and pricing metadata is available. However, the probe did not capture a live 402 challenge (endpoint is POST-only, probe tried HEAD/GET), no response schema or example response is documented, and crawled pages returned only 404 JSON errors. Pricing is clear from the spec. No actual response format is documented.
Warnings
- —Probe did not receive a 402 MPP challenge because HEAD/GET returned 404; the endpoint is POST-only, so liveness could not be fully confirmed via the probe
- —No response schema documented — the output format (e.g., GLB, OBJ, or other 3D format) is not specified in the OpenAPI spec
- —The currency address 0x20c000000000000000000000b9537d11c60e8b50 is assumed to be pathUSD with 6 decimals based on context; if decimals differ, the stated price of $0.046 would be incorrect
Citations
- —Stable Point Aware 3D endpoint is priced at 46,000 base units via Tempo methodhttps://stability-ai.mpp.paywithlocus.com
- —Required field is 'image'; optional fields include texture_resolution, foreground_ratio, guidance_scale, seedhttps://stability-ai.mpp.paywithlocus.com
- —API reference available at platform.stability.aihttps://platform.stability.ai/docs/api-reference
- —LLM-readable docs available at Locushttps://beta.paywithlocus.com/mpp/stability-ai.md