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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 42791 | M1 Pro (16c) | 16 GB | Qwen3.5-9B | 4bit | 1k | 151.4 | 30.4 | 26-03-10 |
| 42792 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 4k | 224.3 | 21.0 | 26-03-10 |
| 42793 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 1k | 222.0 | 21.5 | 26-03-10 |
| 42794 | M4 (10c) | 16 GB | llama-3.1-8b-instruct | 4bit | 4k | 27.2 | 1.6 | 26-03-10 |
| 42795 | M4 (10c) | 16 GB | llama-3.1-8b-instruct | 4bit | 1k | 236.2 | 20.5 | 26-03-10 |
| 42796 | M3 (10c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 309.5 | 33.4 | 26-03-10 |
| 42797 | M4 Pro (16c) | 64 GB | Qwen3.5-9B | 4bit | 4k | 364.7 | 48.0 | 26-03-10 |
| 42798 | M4 Pro (16c) | 64 GB | Qwen3.5-9B | 4bit | 1k | 361.5 | 49.5 | 26-03-10 |
| 42799 | M3 Ultra (60c) | 256 GB | Qwen3.5-35B-A3B | 4bit | 32k | 1,675 | 68.2 | 26-03-10 |
| 42800 | M3 Ultra (60c) | 256 GB | Qwen3.5-35B-A3B | 4bit | 64k | 1,290 | 50.8 | 26-03-10 |