← CanIRunLocal · All hardware · All models
Yes, if the Mac GPU limit is raised. Qwen3.5 122B-A10B (MoE) on Mac Studio M3 Ultra (80c GPU) 96GB at Q4_K_M needs sudo sysctl iogpu.wired_limit_mb=87757 (83.7 GB needed, 72.0 GB usable), about 51 tok/s (estimate).
Hardware: Mac Studio M3 Ultra (80c GPU) 96GB · Model: Qwen3.5 122B-A10B (MoE)
Mac Studio M3 Ultra (80c GPU) 96GB has 96 GB of unified memory and 819 GB/s of memory bandwidth. Mac Studio M3 Ultra (60c GPU) 96GB has the same memory and bandwidth, so the fit numbers match; this page is Mac Studio M3 Ultra (80c GPU) 96GB. The checker leaves macOS a share of unified memory, so Ollama can use 72.0 GB. Qwen3.5 122B-A10B (MoE) (125B, 10B active, Alibaba) at Q4_K_M is a 81 GB download and needs 83.7 GB with headroom for about 8k tokens.
| Best quant | Q4_K_M |
|---|---|
| Download size | 81 GB |
| Memory needed | 83.7 GB |
| Usable fast memory | 72.0 GB |
| Fit | fits if the Mac GPU limit is raised |
| Estimated speed | ~51 tok/s (estimate) |
| Mac GPU limit | sudo sysctl iogpu.wired_limit_mb=87757 |
| Command | ollama run qwen3.5:122b |
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