Senacula Shape Studio
Website: https://senacula.github.io/senacula-shapestudio/ · PyPI: https://pypi.org/project/presidium-shapestudio/ · Releases: https://github.com/senacula/senacula-shapestudio/releases
A free, local image-to-3D studio and MCP server. One picture in, a textured mesh out (GLB
and OBJ with vertex colours, plus a turntable PNG), with permissively licensed models on your
own machine. A desktop app for Windows, macOS and Linux, a web UI, a command line, and an MCP
server for Senacula, Claude, Cursor or any other agent. It runs on the CPU if that is all
there is, and uses an NVIDIA GPU when you have one.
Part of the Senacula family, sibling of Senacula Image Studio,
built on the same architecture and the same rules. Design: Hivemind ADR-051; brand: ADR-065 and 066. Licence:
Apache-2.0.
The policy
Quality is a setting, never a consequence. The default preset is best: the finest
surface grid on the strongest permissively licensed model that fits your machine. The
machine decides how long that takes (cores, GPU, memory), never how good the result is.
Nothing downgrades silently to be faster: the studio estimates the time and tells you.
Faster only with consent. Presets balanced and fast exist and must be chosen
explicitly. Each states its trade-off in words, in the tool description and in
capabilities():
| Preset (TripoSR) | Surface grid | Trade-off |
|---|
best (default) | 256³ | The finest surface the field resolves; most triangles; longest extraction. |
balanced | 192³ | About half the extraction work; slightly softer small details. |
fast | 128³ | A preview-grade surface in a fraction of the time; thin parts may break up. |
The network pass is identical at every preset; only the surface extraction changes. Nothing
else is lowered on its own.
Two tiers, chosen by hardware, never silently. The standard tier is TripoSR (MIT): it
runs on a CPU. The best tier is TRELLIS image-large (MIT): clearly higher fidelity, and it
needs a discrete NVIDIA GPU with 16 GB of memory or more. On a machine without that GPU the
plan says so in fit_note ("Running TripoSR because TRELLIS needs an NVIDIA GPU and this
machine has none. This is a hardware fit, not a speed choice."), and asking for best by
name is refused with the reason instead of being swapped for something else.
Permissive licences only. TripoSR weights and code: MIT. TRELLIS weights: MIT. PyTorch:
BSD. IS-Net and U²-Net background models: Apache-2.0. Excluded by policy, whatever their
quality: Hunyuan3D 2.x (Tencent Hunyuan Non-Commercial License, with territorial
restrictions), Stable Fast 3D, Stable Zero123 and Stable Point Aware 3D (Stability AI
Community License, free only below a revenue threshold).
Offline after setup. The models and, if you want it, the CUDA build of PyTorch download
once, each after explicit consent (--yes or consent=true), into the studio's own data
folder, pinned by SHA-256. After that nothing touches the network; there are no keys or
accounts.
Measured, not promised. Every time estimate comes from runs on your machine (per model,
per device) and is labelled not measured yet until the first one. This README contains no
timing claim this project has not measured. As of 0.1.6 the real TripoSR weights have been
loaded on a user's GPU laptop (the run surfaced two bugs, both fixed in 0.1.6); a complete
measured reconstruction is still to be reported (see Status).
Status, stated plainly
| Part | State |
|---|
| Queue, planner, tiers, consented downloads, UI, CLI, MCP server, GLB/OBJ writer, turntable | Proven by the test suite with a fake engine (no weights needed) |
Vendored TripoSR code (no torchmcubes, transformers, einops, omegaconf) | Runs end to end on random weights; the DINO ViT matches the reference implementation exactly at 224 px and to 1e-2 at 512 px (interpolation recipe); parameter names match the published checkpoint layout |
TripoSR with the real model.ckpt | Loaded on a user's machine (0.1.5, GTX 1660 Ti laptop); the run then failed in the background remover, fixed in 0.1.6. A complete measured reconstruction is still to be reported; the first one writes the machine's profile automatically |
| TRELLIS engine | Not integrated. Catalogue entry, hash pins, fit check and planner tier exist; the engine refuses with the reason until verified on a 16 GB NVIDIA GPU |
| CUDA runtime download for the packaged Windows app | Download and unpack verified by a user (0.1.5). Loading it was broken until 0.1.6 (the launcher never activated it); Settings now shows whether the unpacked build loaded, GPU run still to be confirmed |
| Minimum and recommended requirements per model | Stated in the catalogue and checked against the machine (0.1.6): TripoSR 8 GiB RAM minimum, 16 recommended, NVIDIA 4 GiB optional; TRELLIS NVIDIA 16 GiB minimum, 24 recommended, no CPU path |
| Windows installer, portable zip, MSIX, macOS and Linux builds | Built by the release workflow on GitHub's runners |
| Updates | Off until you say yes (Home asks once). Then one request a day to GitHub's releases API; the Windows installer build can download the verified Setup.exe and install it, the Store updates its own copies, pip installs are told the command (0.1.7) |
以上为 README 的开头部分。其余内容请阅读原文。 查看完整 README ↗
