community pool / model answer
gpt-oss-120b
chat · openai/gpt-oss-120b
70.7 tok/s · model median · 10 runs in the model record.
| Reference rig | Quantization | Median result | Basis | TTFT (ms) | Peak VRAM | Max context | Engines |
|---|---|---|---|---|---|---|---|
| Arc Pro B70 32GB ×2 | 4-bit | 59.4 tok/s | reported | - | 60 GB | 2,048 | llama.cpp |
| GB10 Grace Blackwell 128GB | 4-bit | 34.4 tok/s | reported | - | - GB | 32,768 | vllm |
| Radeon AI Pro R9700 32GB ×3 | 16-bit | 69.9 tok/s | measured | 691 | 68.5 GB | 845 | llama.cpp |
| RTX PRO 6000 Blackwell 96GB | 16-bit | 196 tok/s | measured | 585 | - GB | 15,964 | llama.cpp |
| Ryzen AI Max 395 128GB | 16-bit | 38.8 tok/s | reported | 1,718 | - GB | 845 | 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.