community pool / model answer

Qwen3.8-27B-GSQ-RCO-GGUF

chat · ISTA-DASLab/Qwen3.8-27B-GSQ-RCO-GGUF

33.9 tok/s · model median · 1 runs in the model record.

Answer

56.1tok/s

reportedRTX 3090 24GB3-bit · 1 run · decodebest cell in the pool

  1. RTX 3090 24GB56.1 tok/sreported1 run · peak 23.6 GB · fits 24 GB rig memory
Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
RTX 3090 24GB3-bit56.1 tok/sreported1,07123.6 GB32,768llama.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.