Claude Code: connect to a local model when your quota runs out

Re:Claude Code: connect to a local model when your quota run... by MuffinFlavored on Wednesday February 04, 2026 @09:01PM (#46894694)

Thank you. Could you give a tl;dr on "the full model needs ____ this much VRAM and if you do _____ the most common quantization method it will run in ____ this much VRAM" rough estimate please?

[ Parent | on HN ]

Re:Claude Code: connect to a local model when your quota run... by omneity on Thursday February 05, 2026 @12:07AM (#46895886)

It’s a trivial calculation to make (+/- 10%).

Number of params == “variables” in memory

VRAM footprint ~= number of params * size of a param

A 4B model at 8 bits will result in 4GB vram give or take, same as params. At 4 bits ~= 2GB and so on. Kimi is about 512GB at 4 bits.

[ Parent | on HN ]