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

Can I run Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF on an RTX 3090?

Yes, just

Its Q5_K file (20.2 GB) fits in the 21.6 GB an RTX 3090 can use, with room for ≈ 18K tokens of context.

Full analysis of Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF Everything that runs on RTX 3090 Get the Q5_K file on Hugging Face
ollama run hf.co/HauhauCS/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF:Q5_K

Files and context on RTX 3090

Each file of Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-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 RTX 3090 Context (16-bit)Context (8-bit)
Q8_K31.5 GBDoesn't fit——
Q6_K25.9 GBDoesn't fit——
Q5_K20.2 GBTight18K37K
Q4_K17.9 GBTight52K105K
IQ4_XS15.7 GBFits86K171K
Q3_K13.4 GBFits119K239K
IQ3_M12.8 GBFits129K256K
IQ3_XS12.2 GBFits138K256K
Q2_K10.7 GBFits160K256K
IQ2_M10.3 GBFits166K256K

Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF on other hardware

Popular models that run on RTX 3090

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.