AI tools with open source on CUDA under 12 GB of VRAM
VRAM thresholds, backends and licenses come from the tool fact sheets: every value is collected by code from an official source and stamped with the date it was checked. Requirements marked with an asterisk * are the authors’ own description, not our measurement. Details are on each tool’s page.
Showing 9 of 96
| Tool | Type | Runs | VRAM | Backends | Formats |
|---|---|---|---|---|---|
| MeshAnything | 3D | open source | from 7 GB * | CUDA | — |
| Stable Fast 3Dimage-to-3d | 3D | open source | from 6 GB * | CUDAApple MPS | glb |
| threestudio | 3D | open source | from 6 GB * | CUDA | obj |
| TripoSRimage-to-3d | 3D | open source | from 6 GB * | CUDA | — |
| Barktext-to-speech | Audio | open source | from 12 GB * | CUDA | — |
| MMAudio | Audio | open source | from 6 GB * | CUDA | — |
| Allegrotext-to-video | Video | open source | from 9.3 GB (9.3–27.5 GB) * | CUDA | — |
| FramePack | Video | open source | from 6 GB * | CUDA | — |
| Maestroimage-text-to-video | Video | open source | from 12 GB (12–24 GB) * | CUDA | — |
Requirements not stated by the authors
The official sources for these tools have no data for the selected filter — that means “unknown”, not “won’t work”.
| 3DTopia | 3D | open source | not stated | CUDA | glb |
| CRM | 3D | open source | not stated | — | obj |
| DreamGaussian | 3D | open source | not stated | — | — |
| Era3D | 3D | open source | not stated | — | — |
| Hunyuan3D 2.1image-to-3d | 3D | open source | from 10 GB (10–29 GB) * | — | — |
| InstantMeshimage-to-3d | 3D | open source | not stated | CUDA | obj |
| LGMtext-to-3d | 3D | open source | not stated | — | — |
| LichtFeld-Studio | 3D | open source | not stated | CUDA | ply |
| Michelangelo | 3D | open source | not stated | — | — |
| MVDream | 3D | open source | not stated | — | — |
| Nerfstudio | 3D | open source | not stated | CUDA | — |
| One-2-3-45 | 3D | open source | not stated | — | — |
| OpenLRMimage-to-3d | 3D | open source | not stated | CUDA | obj |
| PhysX-Omniimage-to-3d | 3D | open source | not stated | — | — |
| Point-E | 3D | open source | not stated | — | — |
| Scal3Rimage-to-3d | 3D | open source | not stated | CUDA | — |
| Shap-E | 3D | open source | not stated | — | — |
| TRELLISimage-to-3d | 3D | open source | not stated | CUDA | glb |
| Unique3D | 3D | open source | not stated | CUDA | — |
| video_to_world | 3D | open source | not stated | CUDA | ply |
| Wonder3D | 3D | open source | not stated | — | — |
| Zero123++ | 3D | open source | from 5 GB * | — | — |
| ACE-Steptext-to-audio | Audio | open source | from 8 GB * | — | — |
| ChatTTStext-to-audio | Audio | open source | from 4 GB * | — | — |
| CosyVoicetext-to-speech | Audio | open source | not stated | — | — |
| DiffRhythm | Audio | open source | from 8 GB * | — | — |
| Fish Speechtext-to-speech | Audio | open source | not stated | CUDA | — |
| MOSS-SoundEffecttext-to-audio | Audio | open source | not stated | — | — |
| MusicGentext-to-audio | Audio | open source | not stated | — | — |
| OpenVoicetext-to-speech | Audio | open source | not stated | — | — |
| Stable Audio Opentext-to-audio | Audio | open source | not stated | — | — |
| XTTStext-to-speech | Audio | open source | not stated | CUDA | — |
| YuEtext-generation | Audio | open source | from 24 GB * | — | — |
| AI Comic Builder | Video | open source | not stated | — | — |
| AlayaRendererimage-to-image | Video | open source | not stated | — | — |
| AnimateDiff | Video | open source | from 13 GB * | — | — |
| CogVideoXtext-to-video | Video | open source | not stated | CUDA | — |
| Cosmos3-Super-Image2Videoimage-to-video | Video | open source | not stated | CUDA | — |
| forge-film | Video | open source | not stated | CUDA | — |
| GAE-GeometricAutoEncoderimage-to-video | Video | open source | not stated | — | ply |
| Infinite World | Video | open source | not stated | — | — |
| Latte | Video | open source | not stated | CUDA | — |
| LingBot-Video | Video | open source | not stated | — | — |
| Mochi 1text-to-video | Video | open source | from 80 GB * | — | — |
| nano-world-model | Video | open source | not stated | CUDA | — |
| object-permanencevideo-to-video | Video | open source | not stated | — | — |
| Open-Sora | Video | open source | not stated | — | — |
| Open-Sora-Plan | Video | open source | from 24 GB * | — | — |
| RefAlignimage-text-to-video | Video | open source | not stated | — | — |
| SparkDiffusiontext-to-video | Video | open source | not stated | CUDA | — |
| Stable Video Diffusionimage-to-video | Video | open source | not stated | — | — |
| TE-Speed-MiniMaxH3-OSS | Video | open source | not stated | — | — |
| video-generator-client | Video | open source | not stated | — | — |
| VideoCrafter | Video | open source | not stated | — | — |
| ViMax | Video | open source | not stated | — | — |
| Wantext-to-video | Video | open source | from 8.19 GB * | — | — |
| wind-comic | Video | open source | not stated | — | — |
| WorldCrafterimage-to-video | Video | open source | not stated | CUDA | — |
Fact sheets are re-checked monthly; every value on a tool page has a source link and a check date. The asterisk * means “as described by the author”: a developer’s statement, not an independent measurement.