community pool / model answer
Qwen3.5-2B
chat · Qwen/Qwen3.5-2B
105.9 tok/s · model median · 4 runs in the model record.
| Reference rig | Quantization | Median result | Basis | TTFT (ms) | Peak VRAM | Max context | Engines |
|---|---|---|---|---|---|---|---|
| AMD Ryzen 7 8745H w/ Radeon 780M | 4-bit | 45.9 tok/s | reported | 426 | 2.5 GB | 8,192 | llama.cpp |
| AMD Ryzen 7 8745H w/ Radeon 780M Graphics | 4-bit | 45.9 tok/s | reported | 424 | - GB | 65,536 | llama.cpp |
| RTX 3060 12GB | 4-bit | 166.2 tok/s | reported | 706 | 3.5 GB | 4,608 | llama.cpp |
Context tested
The table above is the tested context for each exact rig and quantization cell. Empty fields remain empty rather than being filled with an estimate.