← CanIRunLocal · All hardware · All models
Yes. Code Llama 7B fits in fast memory on M3 8GB at Q4_K_M (5 GB needed, 5.4 GB usable), about 20 tok/s (estimate).
Hardware: M3 8GB · Model: Code Llama 7B
M3 8GB has 8 GB of unified memory and 100 GB/s of memory bandwidth. M2 8GB has the same memory and bandwidth, so the fit numbers match; this page is M3 8GB. The checker leaves macOS a share of unified memory, so Ollama can use 5.4 GB. Code Llama 7B (6.74B, Meta) at Q4_K_M is a 4.1 GB download and needs 5 GB with headroom for about 8k tokens.
| Best quant | Q4_K_M |
|---|---|
| Download size | 4.1 GB |
| Memory needed | 5 GB |
| Usable fast memory | 5.4 GB |
| Fit | fits in fast memory |
| Estimated speed | ~20 tok/s (estimate) |
| Command | ollama run codellama:7b-instruct-q4_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