← CanIRunLocal · All hardware · All models
Yes. Llama 4 Scout 109B (MoE) fits in fast memory on Mac Studio M3 Ultra (60c GPU) 256GB at Q4_K_M (69.7 GB needed, 192.0 GB usable), about 31 tok/s (estimate).
Hardware: Mac Studio M3 Ultra (60c GPU) 256GB · Model: Llama 4 Scout 109B (MoE)
Mac Studio M3 Ultra (60c GPU) 256GB has 256 GB of unified memory and 819 GB/s of memory bandwidth. Mac Studio M3 Ultra (80c GPU) 256GB has the same memory and bandwidth, so the fit numbers match; this page is Mac Studio M3 Ultra (60c GPU) 256GB. The checker leaves macOS a share of unified memory, so Ollama can use 192.0 GB. Llama 4 Scout 109B (MoE) (109B, 17B active, Meta) at Q4_K_M is a 67 GB download and needs 69.7 GB with headroom for about 8k tokens.
| Best quant | Q4_K_M |
|---|---|
| Download size | 67 GB |
| Memory needed | 69.7 GB |
| Usable fast memory | 192.0 GB |
| Fit | fits in fast memory |
| Estimated speed | ~31 tok/s (estimate) |
| Command | ollama run llama4:16x17b |
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