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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 66231 | M3 Ultra (80c) | 512 GB | GLM-4.7-8bit-gs32 | 8bit | 1k | 221.9 | 13.3 | 26-03-10 |
| 66232 | M3 Ultra (80c) | 512 GB | GLM-4.7-8bit-gs32 | 8bit | 4k | 236.0 | 12.1 | 26-03-10 |
| 66233 | M2 Pro (19c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 4k | 380.1 | 53.1 | 26-03-10 |
| 66234 | M2 Pro (19c) | 32 GB | Qwen3.5-35B-A3B | 4bit | 1k | 370.1 | 55.9 | 26-03-10 |
| 66235 | M4 (10c) | 16 GB | Llama-3.2-3B-Instruct | 4bit | 1k | 547.3 | 46.3 | 26-03-10 |
| 66236 | M4 (10c) | 16 GB | Llama-3.2-3B-Instruct | 4bit | 4k | 507.6 | 38.2 | 26-03-10 |
| 66237 | M4 (10c) | 32 GB | Qwen3.5-0.8B | 4bit | 4k | 1,470 | 112.8 | 26-03-10 |
| 66238 | M4 (10c) | 32 GB | Qwen3.5-0.8B | 4bit | 1k | 1,491 | 127.5 | 26-03-10 |
| 66239 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 4k | 396.3 | 32.4 | 26-03-10 |
| 66240 | M4 (10c) | 16 GB | Qwen3.5-4B | 4bit | 1k | 398.5 | 36.0 | 26-03-10 |