community pool / model answer

Qwen3-14B

chat · Qwen/Qwen3-14B

33.7 tok/s · model median · 20 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
M1 Max 64GB4-bit24.8 tok/sreported51- GB2,048ollama
RTX 3060 12GB4-bit29 tok/smeasured2,5538.7 GB65,536llama.cpp, ollama
RTX 3080 12GB4-bit75 tok/sreported-8.7 GB2,048llama.cpp
RTX 3090 24GB4-bit42.9 tok/smeasured2,6550 GB131,072llama.cpp
RTX3060+RTX2080-pooled-20GB 20GB4-bit35 tok/sreported-8.7 GB2,048ollama

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.