TripoSR
Single-image to textured 3D mesh in under half a second, MIT licensed.
About
TripoSR is a fast 3D reconstruction model released in March 2024 (arXiv 2403.02151), built by Stability AI together with Tripo AI (VAST-AI-Research). It uses an LRM-style transformer architecture, an image encoder, an image-to-triplane decoder, and FlexiCubes mesh extraction, totaling around 1B parameters, to turn a single RGB image into a textured 3D mesh in under 0.5 seconds on an RTX 4090.
It is deliberately lightweight: it runs on consumer GPUs with 6GB or more VRAM, includes built-in background removal, and is released under the MIT license, making it one of the most permissively licensed 3D reconstruction models available.
The honest caveats: it produces vertex colors rather than PBR materials, it is image-only with no text-to-3D mode, and geometry for parts of the object the camera never saw is a plausible hallucination rather than a measurement. For higher fidelity, chain it after an image model or step up to Hunyuan3D or TRELLIS.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 87; Trending score 74/100 from researcher assessment.