Can I run Qwen3.6-27B-GGUF on an NVIDIA H100?

Yes

Its BF16 file (53.8 GB) fits in the 72.0 GB an NVIDIA H100 can use, with room for ≈ 256K tokens of context.

Full analysis of Qwen3.6-27B-GGUF Everything that runs on NVIDIA H100 Get the BF16 file on Hugging Face

llama-server -hf unsloth/Qwen3.6-27B-GGUF:BF16

Files and context on NVIDIA H100

Each file of Qwen3.6-27B-GGUF (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

File	Size	On NVIDIA H100	Context (16-bit)	Context (8-bit)
BF16	53.8 GB	Fits	256K	256K
UD-Q8_K_XL	35.3 GB	Fits	256K	256K
Q8_0	28.6 GB	Fits	256K	256K
UD-Q6_K_XL	25.6 GB	Fits	256K	256K
Q6_K	22.5 GB	Fits	256K	256K
UD-Q5_K_XL	20.0 GB	Fits	256K	256K
Q5_K_M	19.5 GB	Fits	256K	256K
Q5_K_S	19.0 GB	Fits	256K	256K
UD-Q4_K_XL	17.6 GB	Fits	256K	256K
Q4_1	17.3 GB	Fits	256K	256K
Q4_K_M	16.8 GB	Fits	256K	256K
IQ4_NL	16.1 GB	Fits	256K	256K
Q4_K_S	15.9 GB	Fits	256K	256K
Q4_0	15.8 GB	Fits	256K	256K
IQ4_XS	15.4 GB	Fits	256K	256K
UD-Q3_K_XL	14.5 GB	Fits	256K	256K
Q3_K_M	13.6 GB	Fits	256K	256K
Q3_K_S	12.4 GB	Fits	256K	256K
UD-IQ3_XXS	12.0 GB	Fits	256K	256K
UD-Q2_K_XL	11.8 GB	Fits	256K	256K
UD-IQ2_M	10.8 GB	Fits	256K	256K
UD-IQ2_XXS	9.39 GB	Fits	256K	256K

Qwen3.6-27B-GGUF on other hardware

Popular models that run on NVIDIA H100

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.