MicroLLM Lab – Try 7 tiny LLM's in the browser

Re:MicroLLM Lab – Try 7 tiny LLM's in the browser by sourweasel on Monday September 28, 2026 @07:05PM (#49885688)

I would recommend adding the MiniCPM5-1B model. Surprisingly coherent for a 1B model and performs well. On my pixel 9 I get 33 tok/s on CPU. On GPU I get about 26 tok/s but prefill jumps to nearly 500.

[ Parent | on HN ]

There are no comments on this comment yet.