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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 18631 | M5 (10c) | 24 GB | Qwen3.5-4B | 4bit | 1k | 423.7 | 47.6 | 26-03-09 |
| 18632 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 407.4 | 37.9 | 26-03-09 |
| 18633 | M4 (10c) | 16 GB | Qwen3.5-4B-OptiQ | 4bit | 4k | 346.7 | 21.9 | 26-03-09 |
| 18634 | M4 (10c) | 16 GB | Qwen3.5-4B-OptiQ | 4bit | 1k | 328.3 | 23.2 | 26-03-09 |
| 18635 | M4 (10c) | 16 GB | Qwen3.5-0.8B | 8bit | 4k | 1,980 | 102.3 | 26-03-09 |
| 18636 | M4 (10c) | 16 GB | Qwen3.5-0.8B | 8bit | 1k | 2,044 | 109.7 | 26-03-09 |
| 18637 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 400.7 | 36.1 | 26-03-09 |
| 18638 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 4k | 399.5 | 33.9 | 26-03-09 |
| 18639 | M2 Ultra (60c) | 64 GB | Qwen3-Coder-Next | 4bit | 64k | 617.2 | 26.5 | 26-03-09 |
| 18640 | M2 Ultra (60c) | 64 GB | Qwen3-Coder-Next | 4bit | 32k | 737.4 | 39.8 | 26-03-09 |