← CanIRunLocal · All hardware · All models
Yes. Llama 3.3 70B fits in fast memory on M2 Max (38c GPU) 96GB at Q5_K_M (54.1 GB needed, 72.0 GB usable), about 6 tok/s (estimate).
Hardware: M2 Max (38c GPU) 96GB · Model: Llama 3.3 70B
M2 Max (38c GPU) 96GB has 96 GB of unified memory and 400 GB/s of memory bandwidth. The checker leaves macOS a share of unified memory, so Ollama can use 72.0 GB. Llama 3.3 70B (70.6B, Meta) at Q5_K_M is a 50 GB download and needs 54.1 GB with headroom for about 8k tokens.
| Best quant | Q5_K_M |
|---|---|
| Download size | 50 GB |
| Memory needed | 54.1 GB |
| Usable fast memory | 72.0 GB |
| Fit | fits in fast memory |
| Estimated speed | ~6 tok/s (estimate) |
| Command | ollama run llama3.3:70b-instruct-q5_K_M |
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-23.
A project by OnDevice Labs · llms.txt