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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 286141 | M4 (10c) | 16 GB | Meta-Llama-3.1-8B-Instruc... | 4bit | 1k | 237.5 | 22.5 | 26-03-11 |
| 286142 | M1 Pro (14c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 259.1 | 49.2 | 26-03-11 |
| 286143 | M4 Max (40c) | 128 GB | GLM-4.7-Flash | 8bit | 4k | 1,088 | 57.7 | 26-03-11 |
| 286144 | M4 Max (40c) | 128 GB | GLM-4.7-Flash | 8bit | 1k | 1,007 | 63.0 | 26-03-11 |
| 286145 | M1 Pro (14c) | 16 GB | Qwen3.5-4B | 4bit | 4k | 259.0 | 46.5 | 26-03-11 |
| 286146 | M1 Pro (14c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 259.5 | 49.5 | 26-03-11 |
| 286147 | M2 Pro (16c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 16k | 174.4 | 19.8 | 26-03-11 |
| 286148 | M2 Pro (16c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 8k | 192.7 | 27.1 | 26-03-11 |
| 286149 | M2 Pro (16c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 4k | 187.5 | 20.3 | 26-03-11 |
| 286150 | M2 Pro (16c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 1k | 225.6 | 23.9 | 26-03-11 |