community pool / model answer

gemma-4-26B-A4B-it-AWQ-4bit

chat · cyankiwi/gemma-4-26B-A4B-it-AWQ-4bit

369 tok/s · model median · 4 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
RTX 5090 32GB4-bit369 tok/smeasured25- 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.