WHAT NEW IN X11.06 FRI

AI collections from X · 23 collections · 70,791 posts

sonnet-5.5VIDEO

Sonnet 5.5 beats GPT-6.1 Sol in the pelican test

The author says Sonnet 5.5 again beat GPT-6.1 Sol on this video test, though Sol is faster and cheaper and remains his main model for Hermes Agent.

filipe
@filicrovalArchived 2026.10.01Originally posted 2026.09.30
filipe @filicrovalOriginally posted 2026.09.30
Open original on X

yikes, Sonnet 5.5 is also better than GPT-6.1 Sol at this test. 6.1 sol was faster and cheaper tho. it's great for other use cases and still my main model for Hermes Agent, so it's not all bad. Sonnet just keeps winning the pelican test

x.com/filicroval/status/2105297059519541640Archived 2026.10.01 10:28