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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 157031 | M3 Max (40c) | 64 GB | Qwen3-Coder-Next-oQ4 | 4bit | 1k | 969.4 | 61.6 | 26-05-05 |
| 157032 | M3 Max (40c) | 64 GB | Qwen3-Coder-Next-oQ4 | 4bit | 4k | 1,124 | 60.3 | 26-05-05 |
| 157033 | M5 (10c) | 16 GB | Qwen3.5-9B | 4bit | 4k | 243.3 | 17.4 | 26-05-05 |
| 157034 | M5 (10c) | 16 GB | Qwen3.5-9B | 4bit | 1k | 182.0 | 44.5 | 26-05-05 |
| 157035 | M2 Max (38c) | 96 GB | GLM-4.7-Flash | 6bit | 4k | 482.7 | 34.7 | 26-05-05 |
| 157036 | M2 Ultra (60c) | 64 GB | gemma-4-31B-it-OptiQ | 4bit | 4k | 146.8 | 22.7 | 26-05-05 |
| 157037 | M2 Max (38c) | 96 GB | GLM-4.7-Flash | 6bit | 1k | 454.8 | 26.2 | 26-05-05 |
| 157038 | M2 Ultra (60c) | 64 GB | gemma-4-31B-it-OptiQ | 4bit | 1k | 146.3 | 23.9 | 26-05-05 |
| 157039 | M3 Max (40c) | 64 GB | Qwen3-Coder-Next | 4bit | 1k | 962.8 | 64.0 | 26-05-05 |
| 157040 | M3 Max (40c) | 64 GB | Qwen3-Coder-Next | 4bit | 4k | 1,135 | 64.7 | 26-05-05 |