Community Benchmarks
Download oMLX →Real performance numbers from oMLX users around the world. Submit yours from the app (v0.2.6+).
| # | Chip | RAM | Model | Quant | Ctx | PP tok/s | TG tok/s | Date ↓ |
|---|---|---|---|---|---|---|---|---|
| 156801 | M3 Max (40c) | 48 GB | Qwen3.5-35B-A3B | 4bit | 16k | 1,366 | 79.5 | 26-05-04 |
| 156802 | M3 Max (40c) | 48 GB | Qwen3.5-35B-A3B | 4bit | 32k | 1,135 | 72.3 | 26-05-04 |
| 156803 | M3 Max (40c) | 48 GB | Qwen3.5-35B-A3B | 4bit | 4k | 1,463 | 89.3 | 26-05-04 |
| 156804 | M4 Max (40c) | 64 GB | Qwen3.6-27B | 4bit | 4k | 216.0 | 27.2 | 26-05-04 |
| 156805 | M4 Max (40c) | 64 GB | Qwen3.6-27B | 4bit | 1k | 234.0 | 28.6 | 26-05-04 |
| 156806 | M5 Max (40c) | 128 GB | gemma-3-4b-it-abliterated... | 4bit | 1k | 2,702 | 139.8 | 26-05-04 |
| 156807 | M5 Max (40c) | 128 GB | gemma-3-4b-it-abliterated... | 4bit | 4k | 4,802 | 160.5 | 26-05-04 |
| 156808 | M1 Max (32c) | 64 GB | Apertus-8B-2509 | bf16 | 4k | 428.8 | 19.2 | 26-05-04 |
| 156809 | M1 Max (32c) | 64 GB | Apertus-8B-2509 | bf16 | 1k | 434.3 | 20.4 | 26-05-04 |
| 156810 | M4 Pro (16c) | 24 GB | Qwen3.5-9B-OptiQ | 4bit | 4k | 357.8 | 41.7 | 26-05-04 |