Qwen3-30B-A3B-Instruct-2507
6 records across 2 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M5 Max | Qwen3-30B-A3B-Instruct-2507 | 4bit | oMLX | 114.6 | 1364.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| M5 Max | Qwen3-30B-A3B-Instruct-2507 | 4bit | oMLX | 100.3 | 3199.0 | – | – | 4096 | 2026-10-15 | community | – | source |
| M1 Pro | Qwen3-30B-A3B-Instruct-2507 | 4bit | oMLX | 43.2 | 293.6 | – | – | 1024 | 2026-10-16 | community | – | source |
| M1 Pro | Qwen3-30B-A3B-Instruct-2507 | 4bit | oMLX | 35.5 | 313.4 | – | – | 1024 | 2026-10-16 | community | contradiction | source |
| M1 Pro | Qwen3-30B-A3B-Instruct-2507 | 4bit | oMLX | 35.5 | 289.1 | – | – | 4096 | 2026-10-16 | community | – | source |
| M1 Pro | Qwen3-30B-A3B-Instruct-2507 | 4bit | oMLX | 32.9 | 296.0 | – | – | 4096 | 2026-10-16 | community | – | source |