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

Can I run DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF on an NVIDIA B200?

Yes

Its IQ1_M file (121 GB) fits in the 162 GB an NVIDIA B200 can use, with room for an unknown amount of context. That file is under 3 bits per weight, so expect noticeably lower quality.

Full analysis of DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF Everything that runs on NVIDIA B200 Get the IQ1_M file on Hugging Face
llama-server -hf ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF:IQ1_M

Files and context on NVIDIA B200

Each file of DeepSeek-V4.1-Flash-1017-GSQ-RCO-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 B200 Context (16-bit)Context (8-bit)
IQ2_XS155 GBDoesn't fit——
IQ2_XXS138 GBDoesn't fit——
IQ1_M121 GBFitsnot knownnot known
IQ4_NL111 GBFitsnot knownnot known

DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF on other hardware

Popular models that run on NVIDIA B200

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.