DeepSight – give text-only LLMs eyes and hands (zero tokens, on-device) (github.com) 2 points by eliaseffects 2mo ago ↗ HN
1 comment
[ 3.0 ms ] story [ 10.4 ms ] thread