CSM AI explained.
CSM AI is the other company that decided generation alone is not the problem. Its Cube product pairs image and video to 3D with the unglamorous work around it — converting, repairing, remeshing, texturing and managing assets — which makes it a genuine workflow competitor rather than another model wrapper.
What it does well, and where it stops.
- Video-to-3D, which almost nothing else in this list offers
- Repair, remesh and convert tooling around generation, not just generation
- Asset management for teams working at volume
- Public API
- Weaker text-to-3D than the dedicated generators
- No photoreal render studio and no 3D world generation
- No print ordering
Video input is the interesting part
A short video of an object is a far richer input than one photo and much easier to capture than a careful multi-angle set — you walk around the thing with a phone. If your source material is physical objects, that alone can be the deciding feature. Sculpty's multi-view studio asks for 2–5 discrete angles instead, which is more work up front but gives you direct control over which views the model sees.
CSM AI vs Sculpty.
| Feature |  Sculpty | CSM AI |
|---|---|---|
| Text to 3D | ✓ | ◐ |
| Image to 3D | ✓ | ✓ |
| Multi-view to 3D | ✓ | ✓ |
| Multiple AI engines | ✓ | ◐ |
| PBR textures | ✓ | ✓ |
| AI retexturing | ✓ | ✓ |
| Remesh / retopology | ✓ | ✓ |
| Photoreal render studio | ✓ | — |
| Explorable 3D worlds | ✓ | — |
| GLB, STL, OBJ export | ✓ | ✓ |
| Free format converters | ✓ | ✓ |
| Browser 3D viewer | ✓ | ✓ |
| Order a 3D print | ✓ | — |
| Public API | — | ✓ |
| Auto-rigging / animation | — | ◐ |
| Free plan or trial | ✓ | ✓ |
- Multi-view to 3D — Video input effectively covers the multi-angle case.
- Multiple AI engines — Several CSM models rather than a line-up of third-party engines.
What each one costs.
CSM AI does not publish a self-serve price. Sculpty's plans as of 2026-07-26.
CSM AI, answered.
How is CSM different from Meshy or Tripo?+
Its centre of gravity is the workflow rather than the model: convert, repair, remesh and manage assets alongside generation, plus video input. On raw text-to-3D quality the dedicated generators are ahead.
Does CSM do text-to-3D?+
It is primarily image and video driven; text is not its strength. For text prompts, Sculpty runs four engines built for it.
Related pages.
CSM AI facts checked on 2026-07-26.
Five AI 3D engines, one studio.
Generate with five engines, then texture, remesh, render and print — without leaving the app. Free plan, no credit card.
Start free →