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 ↓ |
|---|---|---|---|---|---|---|---|---|
| 19561 | M2 Ultra (76c) | 128 GB | Qwen3.5-122B-A10B | 4bit | 4k | 463.3 | 45.6 | 26-03-08 |
| 19562 | M2 Ultra (76c) | 128 GB | Qwen3.5-122B-A10B | 4bit | 1k | 439.6 | 47.8 | 26-03-08 |
| 19563 | M3 Ultra (80c) | 512 GB | gpt-oss-120b | 8bit | 4k | 631.9 | 54.2 | 26-03-08 |
| 19564 | M3 Ultra (80c) | 512 GB | gpt-oss-120b | 8bit | 1k | 569.3 | 60.7 | 26-03-08 |
| 19565 | M3 Ultra (80c) | 512 GB | Qwen3.5-122B-A10B | 4bit | 1k | 904.5 | 52.5 | 26-03-08 |
| 19566 | M3 Ultra (80c) | 512 GB | Qwen3.5-122B-A10B | 4bit | 4k | 951.6 | 52.0 | 26-03-08 |
| 19567 | M5 (10c) | 24 GB | Qwen3.5-9B-mxfp4 | 4bit | 4k | 214.5 | 25.3 | 26-03-08 |
| 19568 | M5 (10c) | 24 GB | Qwen3.5-9B-mxfp4 | 4bit | 1k | 215.4 | 26.1 | 26-03-08 |
| 19569 | M2 Max (38c) | 96 GB | Qwen3-Coder-Next | 6bit | 32k | 379.4 | 25.5 | 26-03-08 |
| 19570 | M2 Max (38c) | 96 GB | Qwen3-Coder-Next | 6bit | 16k | 412.9 | 38.1 | 26-03-08 |