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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 205951 | M4 (10c) | 32 GB | gemma-4-26b-a4b-it | 4bit | 1k | 363.8 | 32.1 | 26-04-15 |
| 205952 | M1 Max (32c) | 64 GB | Qwen3.5-35B-A3B | 4bit | 4k | 592.4 | 61.9 | 26-04-15 |
| 205953 | M1 Max (32c) | 64 GB | Qwen3.5-35B-A3B | 4bit | 1k | 488.6 | 65.7 | 26-04-15 |
| 205954 | M1 Ultra (48c) | 64 GB | gemma-4-E4B-it | 8bit | 1k | 927.6 | 62.3 | 26-04-15 |
| 205955 | M1 Ultra (48c) | 64 GB | gemma-4-E4B-it | 8bit | 4k | 1,180 | 59.2 | 26-04-15 |
| 205956 | M2 Pro (16c) | 16 GB | Qwen3.5-4B | 4bit | 4k | 317.6 | 56.1 | 26-04-15 |
| 205957 | M2 Pro (16c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 300.6 | 61.0 | 26-04-15 |
| 205958 | M4 Pro (20c) | 64 GB | Qwen3-Coder-Next | 4bit | 4k | 689.8 | 59.3 | 26-04-15 |
| 205959 | M4 Pro (20c) | 64 GB | Qwen3-Coder-Next | 4bit | 1k | 575.0 | 61.8 | 26-04-15 |
| 205960 | M5 (10c) | 16 GB | Qwen3.5-9B-Claude-4.6-Hig... | 8bit | 4k | 632.5 | 14.8 | 26-04-15 |