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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 1 | M1 Pro (16c) | 32 GB | Nex-N2-mini | 4bit | 4k | 310.2 | 49.9 | 26-06-11 |
| 2 | M4 Max (32c) | 36 GB | gemma-4-12B-it-OptiQ | 4bit | 4k | 421.4 | 36.0 | 26-06-11 |
| 3 | M4 Max (32c) | 36 GB | gemma-4-12B-it-OptiQ | 4bit | 1k | 405.1 | 36.5 | 26-06-11 |
| 4 | M4 Max (32c) | 36 GB | Nemotron-Cascade-2-30B-A3... | 4bit | 4k | 1,168 | 141.9 | 26-06-11 |
| 5 | M4 Max (32c) | 36 GB | Nemotron-Cascade-2-30B-A3... | 4bit | 1k | 964.8 | 144.1 | 26-06-11 |
| 6 | M5 Max (40c) | 128 GB | Qwen3.6-27B-UD | 4bit | 64k | 476.1 | 15.5 | 26-06-11 |
| 7 | M5 Max (40c) | 128 GB | Qwen3.6-27B-UD | 4bit | 4k | 807.6 | 20.6 | 26-06-11 |
| 8 | M5 Max (40c) | 128 GB | Qwen3.6-27B-UD | 4bit | 16k | 771.2 | 19.6 | 26-06-11 |
| 9 | M4 (10c) | 24 GB | gemma-4-26b-a4b-it | 4bit | 32k | 307.8 | 8.7 | 26-06-11 |
| 10 | M4 Max (40c) | 128 GB | Qwen3-Coder-30B-A3B-Instr... | 4bit | 16k | 1,008 | 76.0 | 26-06-11 |