logoalt Hacker News

eurekin • yesterday at 2:51 PM • 1 reply • view on HN

> ~1000 bytes per context token per user

Where's that figure comning from? Last time I checked (could be the 3.6 Qwen 27b) single token needed 32kb


Replies

jburgess777 • yesterday at 3:09 PM

The smallest I have seen is DeepSeek 4.1 flash at 890 bytes per token.