Qwen2.5-1.5B
3 records across 2 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Radeon 860M | Qwen2.5-1.5B | Q4_K_M | llama.cpp (RPC) | 50 | – | – | – | – | 2026-09 | community | – | source |
| M1 | Qwen2.5-1.5B | oQ8e | oMLX | 30.1 | 660.4 | – | – | 1024 | 2026-10-13 | community | – | source |
| M1 | Qwen2.5-1.5B | oQ8e | oMLX | 26.9 | 598.0 | – | – | 4096 | 2026-10-13 | community | – | source |