community pool / model answer

gemma-4-19b-a4b-it-REAP-GGUF

chat · mradermacher/gemma-4-19b-a4b-it-REAP-GGUF

77.5 tok/s · model median · 6 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
RTX 3060 12GB ×24-bit77.5 tok/smeasured2,02912.9 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.