Skip to content

Can RTX 3080 10G run Qwen3.5 9B?

Comfortable: GGUF Q6_K needs 8.5 GB at 4K, batch 1; headroom 1.5 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_M6.6ComfortableVerified file
GGUF Q8_010.4TightVerified file
GGUF Q5_K_M7.6ComfortableVerified file
GGUF Q6_K8.5ComfortableVerified file

Context capacity · GGUF Q6_K

ContextTotal GBVerdict
4K8.5Comfortable
16K8.9Tight
32K9.4Tight
128K12.7Does not fit

Generation speed

GGUF Q6_K: 66–88 tok/s · Derived range

Other model / GPU pairs