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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 11 | M1 Max (32c) | 64 GB | Qwen3.6-35B-A3B | 4bit | 4k | 565.7 | 57.6 | 26-06-04 |
| 12 | M4 Pro (20c) | 48 GB | gemma-4-12B-it | 8bit | 1k | 262.4 | 17.5 | 26-06-04 |
| 13 | M4 Pro (20c) | 48 GB | gemma-4-12B-it | 8bit | 4k | 264.3 | 16.0 | 26-06-04 |
| 14 | M2 Max (38c) | 32 GB | gemma-4-26B-A4B-it-OptiQ | 4bit | 4k | 577.3 | 56.5 | 26-06-04 |
| 15 | M2 Max (38c) | 32 GB | gemma-4-26B-A4B-it-OptiQ | 4bit | 1k | 511.7 | 59.0 | 26-06-04 |
| 16 | M4 Max (40c) | 128 GB | gemma-4-12B | 8bit | 4k | 529.3 | 32.0 | 26-06-04 |
| 17 | M4 Max (40c) | 128 GB | gemma-4-12B | 8bit | 1k | 524.6 | 27.0 | 26-06-04 |
| 18 | M4 Max (40c) | 48 GB | Qwen3.5-9B-oQ2-mtp | 2bit | 8k | 854.5 | 98.8 | 26-06-04 |
| 19 | M4 Max (40c) | 48 GB | Qwen3.5-9B-oQ2-mtp | 2bit | 1k | 765.8 | 104.7 | 26-06-04 |
| 20 | M4 Max (40c) | 48 GB | Qwen3.5-9B-oQ2-mtp | 2bit | 4k | 850.3 | 103.6 | 26-06-04 |