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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 66581 | M3 Max (40c) | 64 GB | Qwen3.5-4B | 4bit | 1k | 1,309 | 113.9 | 26-03-11 |
| 66582 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 1k | 223.4 | 20.4 | 26-03-11 |
| 66583 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 4k | 224.6 | 20.2 | 26-03-11 |
| 66584 | M4 (10c) | 16 GB | rnj-1-instruct | 4bit | 4k | 123.2 | 13.3 | 26-03-11 |
| 66585 | M4 (10c) | 16 GB | rnj-1-instruct | 4bit | 1k | 168.7 | 16.5 | 26-03-11 |
| 66586 | M3 Max (40c) | 48 GB | NVIDIA-Nemotron-3-Nano-30... | 4bit | 4k | 1,307 | 119.5 | 26-03-11 |
| 66587 | M3 Max (40c) | 48 GB | NVIDIA-Nemotron-3-Nano-30... | 4bit | 1k | 974.1 | 125.4 | 26-03-11 |
| 66588 | M3 Max (40c) | 48 GB | gpt-oss-20b-MXFP4-Q8 | 4bit | 4k | 932.8 | 82.7 | 26-03-11 |
| 66589 | M3 Max (40c) | 48 GB | gpt-oss-20b-MXFP4-Q8 | 4bit | 1k | 849.9 | 89.7 | 26-03-11 |
| 66590 | M4 (10c) | 16 GB | gemma-3-4b-it | 4bit | 4k | 435.2 | 37.3 | 26-03-11 |