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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 155081 | M3 Max (40c) | 128 GB | Qwen3.6-27B | 8bit | 1k | 209.3 | 11.0 | 26-04-27 |
| 155082 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 128k | 3,168 | 60.5 | 26-04-27 |
| 155083 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 64k | 4,007 | 62.5 | 26-04-27 |
| 155084 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 32k | 4,428 | 66.1 | 26-04-27 |
| 155085 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 16k | 4,291 | 66.7 | 26-04-27 |
| 155086 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 8k | 1,908 | 64.4 | 26-04-27 |
| 155087 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 4k | 1,863 | 66.4 | 26-04-27 |
| 155088 | M5 Pro (20c) | 64 GB | NVIDIA-Nemotron-3-Nano-30... | 8bit | 1k | 1,382 | 67.8 | 26-04-27 |
| 155089 | M4 (10c) | 32 GB | Qwen3.6-35B-A3B | 4bit | 64k | 252.9 | 16.4 | 26-04-27 |
| 155090 | M4 (10c) | 32 GB | Qwen3.6-35B-A3B | 4bit | 1k | 425.5 | 47.7 | 26-04-27 |