GPU comparison
MacBook Air/Pro M4 (16GB) vs MacBook Air/Pro M5 (24GB) for local LLMs
How the two stack up for running open-source LLMs locally — memory, bandwidth, price, and how many of the 34 tracked models each one can run.
| MacBook Air/Pro M4 (16GB) | MacBook Air/Pro M5 (24GB) | |
|---|---|---|
| Memory | 16.0 GB | 24.0 GB |
| Bandwidth | 120 GB/s | 153 GB/s |
| Price (approx) | $1,199 | $1,299 |
| LLMs it runs | 13 of 34 | 21 of 34 |
| Best model it runs | Devstral Small 2 24B · 8–12 tok/s | Gemma 4 31B · 6–9 tok/s |
The MacBook Air/Pro M5 (24GB) runs 8 more of the tracked models (21 vs 13), thanks to its 24.0 GB of memory. The MacBook Air/Pro M5 (24GB) has more memory bandwidth (153 GB/s), so it generates tokens faster at the same model and quant.