GLM-4.6
6 records across 1 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GB10 | GLM-4.6 | IQ2_XXS+Q5_K | llama.cpp (CUDA) | 17.14 | – | – | – | 113664 | 2026-10-03 | community | – | source |
| GB10 | GLM-4.6 | IQ2_XXS+Q5_K | llama.cpp (CUDA) | 14.36 | – | – | – | 32768 | 2026-10-03 | community | – | source |
| GB10 | GLM-4.6 | IQ2_XXS+Q5_K | llama.cpp (CUDA) | 12.17 | – | – | – | 32768 | 2026-10-03 | community | – | source |
| GB10 | GLM-4.6 | IQ2_XXS+Q5_K | llama.cpp (CUDA) | 11.45 | – | – | – | 32768 | 2026-10-03 | community | – | source |
| GB10 | GLM-4.6 | IQ2_XXS+Q5_K | llama.cpp (CUDA) | 4.10 | 197 | – | – | 57600 | 2026-10-03 | community | – | source |
| GB10 | GLM-4.6 | IQ2_XXS+Q5_K | llama.cpp (CUDA) | 3.61 | 125 | – | – | 111900 | 2026-10-03 | community | – | source |