community pool / model answer

Qwen3.5-9B-MTP-GGUF

chat · unsloth/Qwen3.5-9B-MTP-GGUF

70.6 tok/s · model median · 5 runs in the model record.

Reference rig	Quantization	Median result	Basis	TTFT (ms)	Peak VRAM	Max context	Engines
M5 Pro 24GB	4-bit	45.3 tok/s	reported	532	- GB	4,224	llama.cpp
RTX 3060 12GB ×2	4-bit	47.3 tok/s	reported	-	9.7 GB	128,000	llama.cpp
RTX 3090 24GB ×2	8-bit	70.6 tok/s	reported	1,051	- GB	262,144	llama.cpp
RX 7900 XTX 24GB	6-bit	154.4 tok/s	reported	-	16.2 GB	262,144	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.