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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 279741 | M3 Max (40c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 4bit | 4k | 1,289 | 118.9 | 26-03-17 |
| 279742 | M3 Max (40c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 4k | 1,253 | 78.9 | 26-03-17 |
| 279743 | M3 Max (40c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 1k | 1,062 | 81.6 | 26-03-17 |
| 279744 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 4k | 211.7 | 20.6 | 26-03-17 |
| 279745 | M3 Max (40c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 4bit | 64k | 978.4 | 71.5 | 26-03-17 |
| 279746 | M3 Max (40c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 4bit | 1k | 1,094 | 124.3 | 26-03-17 |
| 279747 | M3 Max (40c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 4bit | 4k | 1,284 | 118.9 | 26-03-17 |
| 279748 | M2 Max (30c) | 32 GB | Qwen3.5-27B-Claude-4.6-Op... | 4bit | 4k | 92.4 | 19.3 | 26-03-17 |
| 279749 | M2 Max (30c) | 32 GB | Qwen3.5-27B-Claude-4.6-Op... | 4bit | 1k | 90.8 | 20.4 | 26-03-17 |
| 279750 | M3 Ultra (60c) | 96 GB | Qwen3.5-27B | 8bit | 4k | 321.0 | 21.9 | 26-03-17 |