community pool / model answer

Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive

chat · HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive

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

Reference rig	Quantization	Median result	Basis	TTFT (ms)	Peak VRAM	Max context	Engines
M5 Max 128GB	6-bit	99.7 tok/s	reported	20	- GB	8,192	llama.cpp
Radeon AI Pro R9700 32GB	4-bit	126.6 tok/s	reported	-	- GB	2,048	llama.cpp
Radeon AI Pro R9700 32GB ×3	8-bit	76.3 tok/s	reported	354	49 GB	787	llama.cpp
Ryzen AI Max 395 128GB	4-bit	64.9 tok/s	reported	201	24.2 GB	32,768	llama.cpp
T4 16GB	2-bit	59.3 tok/s	reported	-	- GB	2,048	llama.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.