Model Anatomy See what's inside open-weight AI models

Can I run Qwen3.8-27B-AP-GGUF on an NVIDIA H100?

Yes

Its Q4_K_XL file (17.6 GB) fits in the 72.0 GB an NVIDIA H100 can use, with room for ≈ 256K tokens of context.

Full analysis of Qwen3.8-27B-AP-GGUF Everything that runs on NVIDIA H100 Get the Q4_K_XL file on Hugging Face
ollama run hf.co/agentionai/Qwen3.8-27B-AP-GGUF:Q4_K_XL

Files and context on NVIDIA H100

Each file of Qwen3.8-27B-AP-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.

FileSizeOn NVIDIA H100 Context (16-bit)Context (8-bit)
Q4_K_XL17.6 GBFits256K256K
Q4_K_M16.5 GBFits256K256K
IQ4_XS14.3 GBFits256K256K
Q3_K_XL13.1 GBFits256K256K
IQ3_S12.0 GBFits256K256K
IQ3_XS11.5 GBFits256K256K
IQ3_XXS10.7 GBFits256K256K
IQ2_S9.61 GBFits256K256K

Qwen3.8-27B-AP-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.