Community Benchmarks
Download oMLX →Real performance numbers from oMLX users around the world. Submit yours from the app (v0.2.6+).
Reset filters| # | Chip | RAM | Model | Quant | Ctx | PP tok/s | TG tok/s | Date ↓ |
|---|---|---|---|---|---|---|---|---|
| 441 | M5 Pro (20c) | 48 GB | Qwen3.6-35B-A3B | 4bit | 64k | 1,294 | 51.9 | 26-07-16 |
| 442 | M3 Max (30c) | 96 GB | Qwen3.6-35B-A3B-oQ8-fp16-... | 8bit | 64k | 465.3 | 27.4 | 26-07-16 |
| 443 | M4 Max (40c) | 128 GB | Qwen3-Coder-Next-oQ8e | 8bit | 64k | 531.0 | 22.2 | 26-07-16 |
| 444 | M5 Pro (16c) | 24 GB | Qwen3.5-4B | 4bit | 64k | 583.5 | 49.2 | 26-07-16 |
| 445 | M2 Pro (16c) | 32 GB | Qwen3.6-35B-A3B | 4bit | 64k | 164.3 | 30.2 | 26-07-16 |
| 446 | M2 Max (30c) | 32 GB | Qwen3.5-9B | 4bit | 64k | 240.1 | 38.6 | 26-07-16 |
| 447 | M3 Max (40c) | 128 GB | Ternary-Bonsai-27B | 2bit | 64k | 120.2 | 12.3 | 26-07-16 |
| 448 | M3 Max (30c) | 96 GB | Qwen3.5-122B-A10B | 4bit | 64k | 188.5 | 17.2 | 26-07-16 |
| 449 | M2 Max (30c) | 32 GB | Qwen3.5-2B | 4bit | 64k | 894.6 | 86.4 | 26-07-16 |
| 450 | M2 Max (30c) | 32 GB | Qwopus3.5-9B-v3 | 8bit | 64k | 223.7 | 28.3 | 26-07-16 |