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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 206161 | M3 Ultra (60c) | 96 GB | Qwen3.5-35B-A3B | 4bit | 16k | 2,011 | 83.4 | 26-04-15 |
| 206162 | M3 Ultra (60c) | 96 GB | Qwen3.5-35B-A3B | 4bit | 32k | 1,745 | 72.2 | 26-04-15 |
| 206163 | M3 Ultra (60c) | 96 GB | Qwen3.5-35B-A3B | 4bit | 8k | 2,132 | 92.5 | 26-04-15 |
| 206164 | M3 Ultra (60c) | 96 GB | Qwen3.5-35B-A3B | 4bit | 1k | 1,809 | 398.3 | 26-04-15 |
| 206165 | M3 Ultra (60c) | 96 GB | Qwen3.5-35B-A3B | 4bit | 4k | 1,966 | 100.2 | 26-04-15 |
| 206166 | M2 Max (30c) | 32 GB | QwenPaw-Flash-9B-oQ4 | 4bit | 32k | 296.7 | 38.7 | 26-04-15 |
| 206167 | M2 Max (30c) | 32 GB | QwenPaw-Flash-9B-oQ4 | 4bit | 16k | 304.3 | 45.9 | 26-04-15 |
| 206168 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 4k | 223.2 | 21.0 | 26-04-15 |
| 206169 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 1k | 217.0 | 21.6 | 26-04-15 |
| 206170 | M2 Max (30c) | 32 GB | QwenPaw-Flash-9B-oQ6 | 6bit | 32k | 296.6 | 31.4 | 26-04-15 |