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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 4071 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 1k | 222.0 | 20.6 | 26-03-09 |
| 4072 | M2 Ultra (60c) | 64 GB | Qwen3.5-27B | 4bit | 32k | 169.1 | 20.5 | 26-03-09 |
| 4073 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 4k | 223.2 | 20.1 | 26-03-09 |
| 4074 | M4 (10c) | 24 GB | Qwen3.5-9B | 4bit | 1k | 222.1 | 20.7 | 26-03-09 |
| 4075 | M4 Pro (20c) | 64 GB | GLM-4.7-Flash | 6bit | 4k | 546.4 | 42.5 | 26-03-09 |
| 4076 | M4 Pro (20c) | 64 GB | GLM-4.7-Flash | 6bit | 1k | 608.4 | 49.8 | 26-03-09 |
| 4077 | M1 Max (32c) | 64 GB | Qwen3.5-9B | 4bit | 4k | 311.9 | 53.2 | 26-03-09 |
| 4078 | M1 Max (32c) | 64 GB | Qwen3.5-9B | 4bit | 1k | 308.2 | 55.1 | 26-03-09 |
| 4079 | M4 Pro (20c) | 64 GB | Qwen3-Coder-Next | 4bit | 4k | 660.7 | 60.6 | 26-03-09 |
| 4080 | M4 Pro (20c) | 64 GB | Qwen3-Coder-Next | 4bit | 1k | 574.9 | 64.4 | 26-03-09 |