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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 13401 | M1 Max (24c) | 64 GB | Qwen3.5-9B | 8bit | 1k | 238.4 | 35.8 | 26-03-09 |
| 13402 | M1 Max (24c) | 64 GB | Qwen3.5-9B | 8bit | 4k | 239.7 | 34.0 | 26-03-09 |
| 13403 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 405.9 | 39.1 | 26-03-09 |
| 13404 | M2 Ultra (60c) | 64 GB | Qwen3-Coder-Next | 4bit | 4k | 789.3 | 58.9 | 26-03-09 |
| 13405 | M2 Ultra (60c) | 64 GB | Qwen3-Coder-Next | 4bit | 1k | 555.6 | 61.8 | 26-03-09 |
| 13406 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 4k | 225.7 | 20.4 | 26-03-09 |
| 13407 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 1k | 225.0 | 20.8 | 26-03-09 |
| 13408 | M1 Max (24c) | 64 GB | gpt-oss-20b | 8bit | 4k | 276.0 | 46.2 | 26-03-09 |
| 13409 | M1 Max (24c) | 64 GB | gpt-oss-20b | 8bit | 1k | 246.3 | 50.8 | 26-03-09 |
| 13410 | M3 Pro (18c) | 36 GB | Qwen3.5-35B-A3B | 4bit | 16k | 609.9 | 37.5 | 26-03-09 |