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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 65121 | M3 Ultra (80c) | 512 GB | GLM-4.7-8bit-gs32 | 8bit | 32k | 134.5 | 6.8 | 26-03-10 |
| 65122 | M2 Ultra (60c) | 128 GB | Qwen3.5-35B-A3B | 8bit | 4k | 1,118 | 64.6 | 26-03-10 |
| 65123 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 8k | 222.3 | 20.1 | 26-03-10 |
| 65124 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 4k | 225.7 | 21.0 | 26-03-10 |
| 65125 | M2 Ultra (60c) | 128 GB | Qwen3.5-35B-A3B | 8bit | 1k | 943.0 | 67.3 | 26-03-10 |
| 65126 | M4 (10c) | 32 GB | Qwen3.5-9B | 4bit | 4k | 186.1 | 19.7 | 26-03-10 |
| 65127 | M4 (10c) | 32 GB | Qwen3.5-9B | 4bit | 1k | 184.3 | 18.4 | 26-03-10 |
| 65128 | M4 (10c) | 16 GB | Qwen3.5-0.8B | unknown | 4k | 1,535 | 12.4 | 26-03-10 |
| 65129 | M4 (10c) | 16 GB | Qwen3.5-0.8B | unknown | 1k | 1,349 | 13.2 | 26-03-10 |
| 65130 | M4 (10c) | 16 GB | Qwen3-VL-Embedding-2B | unknown | 4k | 767.3 | 22.6 | 26-03-10 |