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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 201 | M3 Ultra (60c) | 256 GB | gemma-4-31b-it | 8bit | 64k | 193.3 | 11.4 | 26-05-27 |
| 202 | M4 Max (40c) | 128 GB | gemma-4-26b-a4b-it | 8bit | 64k | 962.4 | 43.1 | 26-05-27 |
| 203 | M4 Pro (16c) | 48 GB | gemma-4-26B-A4B-it | 4bit | 64k | 369.5 | 37.2 | 26-05-27 |
| 204 | M3 Ultra (60c) | 256 GB | gemma-4-31b-it | 8bit | 64k | 193.2 | 12.1 | 26-05-27 |
| 205 | M5 Pro (20c) | 64 GB | gemma-4-26b-a4b-it | 8bit | 64k | 1,318 | 29.0 | 26-05-27 |
| 206 | M4 Max (40c) | 128 GB | gemma-4-26B-A4B-it-oQ8 | 8bit | 64k | 1,015 | 49.9 | 26-05-27 |
| 207 | M4 Pro (16c) | 24 GB | gemma-4-E4B-TurboQuant | 4bit | 64k | 756.2 | 27.6 | 26-05-27 |
| 208 | M4 Max (40c) | 128 GB | gemma-4-26b-a4b-it | 8bit | 64k | 954.5 | 46.6 | 26-05-27 |
| 209 | M4 Pro (16c) | 24 GB | gemma-4-E4B-TurboQuant | 4bit | 64k | 759.3 | 29.4 | 26-05-27 |
| 210 | M4 Pro (16c) | 24 GB | gemma-4-E4B-it | 4bit | 64k | 762.5 | 24.3 | 26-05-27 |