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

Can I run DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic on an RTX 5070?

Yes, just

Its published weights (9.32 GB) fit in the 10.8 GB an RTX 5070 can use, with room for ≈ 43K tokens of context.

Full analysis of DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic Everything that runs on RTX 5070 Get the weights on Hugging Face
hf download xtools-at/DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic

Files and context on RTX 5070

Each file of DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic (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 5070 Context (16-bit)Context (8-bit)
As published9.32 GBTight43K85K
4-bit (estimate)2.62 GBFits242K256K

DeepSeek-V4-Pro-Qwen3.5-4B-UNCENSORED-heretic on other hardware

Popular models that run on RTX 5070

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.