What can I run on Apple Silicon 32 GB?
32 GB unified memoryApple M-seriesView matrix for this target →
A memory class, not a single chip: any M-series Mac with 32 GB unified memory (e.g. M4 configured to 32GB in the 2024 Mac mini). Chip generation (M1–M5) and variant (base/Pro/Max/Ultra, GPU core count) change inference speed substantially, so runs on this target must record the exact chip in platform.chip. PyTorch uses the MPS backend; unified memory is shared with macOS, so usable model memory is lower than 32 GB.
Policy by policy
| Policy | Params | Inference memory (sourced, with precision) | On Apple Silicon 32 GB | Best measured |
|---|---|---|---|---|
| ACT Imitation | 80M | 211 MB measured fp32 | no data | |
| Diffusion Policy Imitation | — | 1.12 GB measured fp32 | no data | |
| EO-1 VLA | 3B | ≥6.5 GB | no data | |
| Evo-1 VLA | 770M | — | no data | |
| Fast-WAM World model | 6B | — | no data | |
| GR00T N1.7 VLA | 3B | ≥16 GB | no data | |
| HIL-SERL RL | — | — | no data | |
| LingBot-VA World model | 5.3B | ≥18 GB | no data | |
| MolmoAct2 VLA | — | 12.1 GB measured bf16 | no data | |
| Multitask DiT Imitation | 450M | — | no data | |
| OpenVLA VLA | 7.5B | 15 GB measured bf16 | no data | |
| OpenVLA-OFT VLA | — | ≥16 GB | no data | |
| SmolVLA VLA | 450M | 1.75 GB measured fp32 | no data | |
| TD-MPC RL | — | — | no data | |
| VLA-JEPA VLA | — | — | no data | |
| VQ-BeT Imitation | — | — | no data | |
| WALL-OSS VLA | — | — | no data | |
| X-VLA VLA | 900M | — | no data | |
| π0 VLA | 3.3B | 13.32 GB measured fp32 | no data | |
| π0-FAST VLA | — | ≥8 GB | no data | |
| π0.5 VLA | — | ≥8 GB | no data |
“Exceeds memory” compares a policy's upstream-stated minimum inference memory (never a measured peak, which is precision-specific) with this target's unified (shared CPU/GPU) memory, so the real budget is lower. It is arithmetic, not observation, and is labelled inferred. Quantised or split runtimes can change the answer; a measured run always wins.
All measured runs on Apple Silicon 32 GB
No measured runs on Apple Silicon 32 GB yet. A run needs runtime, precision, camera count. Contribute one.
Sources
- T1 · upstream documentedvendor-docsverified 2026-10-24 (1d ago)
Apple tech specs for Mac mini (2024): M4 ships with 16GB (configurable to 24GB or 32GB); M4 Pro with 24GB (configurable to 48GB or 64GB) unified memory.
“Apple M4 chip 16GB unified memory Configurable to: 24GB or 32GB Apple M4 Pro chip 24GB unified memory Configurable to: 48GB or 64GB”
https://support.apple.com/en-us/121555 - T1 · upstream documentedofficial-docssource 2026-08-03verified 2026-10-24 (1d ago)lerobot 0.6.1
LeRobot docs: Apple Silicon GPUs are used through the PyTorch MPS backend (--policy.device=mps).
“- **MPS**: acceleration with Apple Silicon GPUs”
https://github.com/huggingface/lerobot/blob/v0.6.1/docs/source/torch_accelerators.mdx