Rent a GPU

Qwen3.6 27B hardware requirements

27.8B parameters. Needs about 18 GB at Q4_K_M with 8K context.

Parameters28BDense
Q4_K_M · 8K18 GBthe usual download
FP16 · 8K56.7 GBfull precision
Runs well on18/51GPUs and Macs we track

Memory needed at each quantization

Weights, KV cache and runtime overhead, at 8K and 32K tokens of context.

QuantQuality8K context32K context
FP16Full quality56.7 GB58.3 GB
Q8_0Near-lossless30.7 GB32.3 GB
Q6_KExcellent23.9 GB25.5 GB
Q5_K_MVery good20.9 GB22.5 GB
Q4_K_MGood, most popular18 GB19.6 GB
Q3_K_MNoticeably degraded14.7 GB16.3 GB
Q2_KHeavily degraded11.6 GB13.2 GB

Run it with Ollama

Downloads Q4_K_M by default.

ollama run qwen3.6:27b

Model card

Qwen/Qwen3.6-27B on Hugging Face: licence, prompt format and GGUF builds.

Run it in the cloud

Run Qwen3.6 27B on a rented GPU today.

It needs about 18 GB at Q4_K_M: a rented RTX 3090 / RTX 4090 runs it for about $0.18 to $0.89 an hour.

RunPodRent a GPU on RunPodVast.aiRent a GPU on Vast.ai

Every GPU and Mac checked

Sorted by memory. Bars show how much of each card Qwen3.6 27B fills at the best quant that fits.

18 Runs well

NameMemory usedBest quantSpeed

22 Fits with heavy quantization

NameMemory usedBest quantSpeed

11 Too big

NameMemory usedBest quantSpeed

Other Qwen models