community pool / model answer

Qwen3-30B-A3B

chat · Qwen/Qwen3-30B-A3B

90.5 tok/s · model median · 8 runs in the model record.

Reference rigQuantizationMedian resultBasisTTFT (ms)Peak VRAMMax contextEngines
M1 Max 64GB4-bit62.6 tok/sreported24- GB2,048ollama
M3 Ultra 512GB4-bit102 tok/sreported76- GB8,192llama.cpp
RTX 2080 8GB4-bit19 tok/sreported-12 GB2,048ollama
RTX 3060 12GB4-bit107 tok/smeasured-12 GB2,048llama.cpp, ollama
RTX 3080 12GB4-bit31 tok/sreported-8.2 GB2,048ollama
RTX3060+RTX2080-pooled-20GB 20GB4-bit91 tok/sreported-12 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.