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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 161421 | M5 Max (40c) | 128 GB | Qwen3.6-27B | 6bit | 1k | 730.5 | 24.5 | 26-04-25 |
| 161422 | M5 Max (40c) | 128 GB | gemma-4-31b-it-mxfp8 | 8bit | 4k | 649.9 | 14.8 | 26-04-25 |
| 161423 | M5 Max (40c) | 128 GB | gemma-4-31b-it-mxfp8 | 8bit | 1k | 606.8 | 15.2 | 26-04-25 |
| 161424 | M4 Max (40c) | 64 GB | Qwen3.6-35B-A3B-UD | 4bit | 4k | 1,673 | 96.9 | 26-04-25 |
| 161425 | M4 Max (40c) | 64 GB | Qwen3.6-35B-A3B-UD | 4bit | 1k | 1,128 | 98.6 | 26-04-25 |
| 161426 | M1 Max (32c) | 64 GB | Qwen3-Coder-Next | 4bit | 4k | 450.1 | 45.8 | 26-04-25 |
| 161427 | M1 Max (32c) | 64 GB | Qwen3-Coder-Next | 4bit | 1k | 381.6 | 48.1 | 26-04-25 |
| 161428 | M3 Ultra (60c) | 96 GB | gpt-oss-20b-MXFP4-Q8 | 4bit | 1k | 1,129 | 112.4 | 26-04-25 |
| 161429 | M3 Ultra (60c) | 96 GB | gpt-oss-20b-MXFP4-Q8 | 4bit | 4k | 1,748 | 103.5 | 26-04-25 |
| 161430 | M1 Max (32c) | 64 GB | Qwen3.5-27B-unsloth | 3bit | 4k | 92.5 | 14.4 | 26-04-25 |