community pool / model answer

Qwen3.6-35B-A3B-NVFP4

chat · RedHatAI/Qwen3.6-35B-A3B-NVFP4

102.1 tok/s · model median · 5 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
GB10 Grace Blackwell 128GB4-bit94.8 tok/smeasured15387 GB262,144vllm
RTX 5090 32GB4-bit183.3 tok/sreported12231.3 GB262,144vllm
RTX 5090 32GB ×24-bit340.8 tok/sreported1,11158.9 GB32,768vllm

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.