community pool / model answer

Qwen3.6-35B-A3B-GPTQ-Int4

chat · palmfuture/Qwen3.6-35B-A3B-GPTQ-Int4

80.9 tok/s · model median · 23 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
Arc Pro B70 32GB4-bit248.3 tok/sreported-- GB32,768vllm
RTX 5090 32GB4-bit80.5 tok/smeasured-- GB2,048vllm

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.