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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 331641 | M3 Ultra (80c) | 512 GB | GLM-4.7-8bit-gs32 | 8bit | 4k | 234.9 | 12.2 | 26-03-10 |
| 331642 | M3 Ultra (80c) | 512 GB | GLM-4.7-8bit-gs32 | 8bit | 1k | 220.4 | 13.4 | 26-03-10 |
| 331643 | M4 Pro (16c) | 24 GB | Llama-3.2-3B-Instruct | 8bit | 4k | 795.9 | 52.0 | 26-03-10 |
| 331644 | M4 Pro (16c) | 24 GB | Llama-3.2-3B-Instruct | 8bit | 1k | 818.6 | 60.0 | 26-03-10 |
| 331645 | M4 (10c) | 16 GB | Qwen3.5-9B | 4bit | 1k | 224.3 | 21.1 | 26-03-10 |
| 331646 | M3 Max (40c) | 64 GB | Qwen3.5-27B | 4bit | 1k | 214.8 | 22.8 | 26-03-10 |
| 331647 | M3 Max (40c) | 64 GB | Qwen3.5-27B | 4bit | 4k | 215.9 | 21.6 | 26-03-10 |
| 331648 | M5 (10c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 4k | 490.6 | 59.3 | 26-03-10 |
| 331649 | M5 (10c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 1k | 479.5 | 61.6 | 26-03-10 |
| 331650 | M4 Pro (20c) | 64 GB | Qwen3.5-35B-A3B | 8bit | 64k | 487.2 | 21.4 | 26-03-10 |