Community Benchmarks
Download oMLX →Real performance numbers from oMLX users around the world. Submit yours from the app (v0.2.6+).
Reset filters| # | Chip | RAM | Model | Quant | Ctx | PP tok/s | TG tok/s | Date ↓ |
|---|---|---|---|---|---|---|---|---|
| 21 | M4 Max (40c) | 64 GB | Qwen3.6-35B-A3B | 8bit | 64k | 832.9 | 60.6 | 26-07-22 |
| 22 | M3 Ultra (60c) | 256 GB | gemma-4-31B-it-qat-OptiQ | 4bit | 64k | 181.1 | 15.7 | 26-07-22 |
| 23 | M4 Max (32c) | 36 GB | gemma-4-e4b-it | 4bit | 64k | 1,283 | 47.2 | 26-07-22 |
| 24 | M3 Ultra (60c) | 256 GB | ThinkingCap-Qwen3.6-27B-O... | 4bit | 64k | 256.7 | 24.7 | 26-07-22 |
| 25 | M5 Max (40c) | 128 GB | Laguna-S-2.1-NVFP4 | 4bit | 64k | 894.9 | 32.3 | 26-07-22 |
| 26 | M3 Ultra (60c) | 256 GB | Laguna-XS-2.1 | 8bit | 64k | 790.3 | 59.7 | 26-07-22 |
| 27 | M5 Max (40c) | 128 GB | Laguna-S-2.1-oQ5e | 5bit | 64k | 658.8 | 40.2 | 26-07-22 |
| 28 | M2 Ultra (60c) | 64 GB | Qwen3.5-35B-A3B-OptiQ | 4bit | 64k | 795.1 | 55.0 | 26-07-22 |
| 29 | M2 Ultra (60c) | 64 GB | Qwen3.6-35B-A3B-Claude-4.... | 4bit | 64k | 1,159 | 63.5 | 26-07-22 |
| 30 | M4 Pro (20c) | 48 GB | gpt-oss-20b-MXFP4-Q8 | 4bit | 64k | 467.1 | 34.5 | 26-07-22 |