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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 336551 | M3 Ultra (80c) | 512 GB | GLM-4___7-Flash | 8bit | 1k | 1,590 | 54.4 | 26-03-12 |
| 336552 | M3 Ultra (80c) | 512 GB | GLM-4___7-Flash | 8bit | 4k | 1,662 | 52.0 | 26-03-12 |
| 336553 | M1 (8c) | 16 GB | Qwen3.5-9B | 4bit | 4k | 77.7 | 11.9 | 26-03-12 |
| 336554 | M1 (8c) | 16 GB | Qwen3.5-9B | 4bit | 1k | 77.2 | 12.6 | 26-03-12 |
| 336555 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 4k | 404.8 | 37.0 | 26-03-12 |
| 336556 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 405.9 | 38.7 | 26-03-12 |
| 336557 | M1 Ultra (48c) | 64 GB | Qwen3-Embedding-0.6B-4bit... | 4bit | 4k | 3,212 | 149.3 | 26-03-12 |
| 336558 | M1 Ultra (48c) | 64 GB | Qwen3-Embedding-0.6B-4bit... | 4bit | 1k | 2,192 | 205.6 | 26-03-12 |
| 336559 | M1 Ultra (64c) | 64 GB | Qwen3.5-35B-A3B-Hereticmx... | 8bit | 32k | 942.4 | 33.6 | 26-03-12 |
| 336560 | M1 Ultra (64c) | 64 GB | Qwen3.5-35B-A3B-Hereticmx... | 8bit | 64k | 712.3 | 22.4 | 26-03-12 |