Skip to content

Can Radeon AI PRO R9700 32G run Qwen3.6 27B?

Comfortable: GGUF Q5_K_M needs 22.2 GB at 4K, batch 1; headroom 9.8 GB.

Text input, FP16 KV cache, one card. Unified-memory capacity uses the GPU guide’s usable limit; negative headroom may require host offload. Speed source settings may differ from this 4K estimate.

Measured runs

No measured run for this pair. Sizing is an estimate, not a runtime test.

Quant memory at 4K

QuantTotal GBVerdictWeight sizing
GGUF Q4_K_M19.1ComfortableVerified file
GGUF Q5_K_M22.2ComfortableVerified file
GGUF Q8_030.7TightVerified file

Context capacity · GGUF Q5_K_M

ContextTotal GBVerdict
4K22.2Comfortable
16K23.0Comfortable
32K24.1Comfortable
128K30.7Tight

Generation speed

GGUF Q5_K_M: 17–22 tok/s · Derived range

Other model / GPU pairs