community pool / model answer

Agents-A1

chat · InternScience/Agents-A1

76.7 tok/s · model median · 12 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Arc Pro B70 32GB4-bit73.2 tok/sreported77- GB262,144llama.cpp
RTX 3060 12GB ×24-bit91.7 tok/smeasured3,12620.9 GB65,536llama.cpp
RX 570 4GB ×24-bit13.6 tok/smeasured188- GB2,048llama.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.