WHAT NEW IN X10.29 THU

AI collections from X · 22 collections · 58,925 posts

gpt-6VIDEO

Meta Muse Wonder tops GPT-6 Astra on agent benchmarks

A video lists scores comparing Meta Muse Wonder with Opus 5.5 and GPT-6 Astra on three agent tests: SWE-Marathon, AutomationBench and MemoryBench.

Aditya Jhajharia
@AdityaJhajhar12Archived 15 days agoOriginally posted 2026.10.13
Aditya Jhajharia @AdityaJhajhar12Originally posted 2026.10.13
Open original on X

current ai situation after Meta Muse Wonder dropped💀 Beat Opus 5.5 and GPT-6 Astra on agent work 😳 - SWE-Marathon: 57.5 vs 46.3 opus - AutomationBench: 45.4 vs 41.6 gpt-6 - MemoryBench: 85.5 vs 81.5 opus

This post quotes another post: x.com/finkd/status/2109689305073614918

x.com/adityajhajhar12/status/2109709361706938810Archived 2026.10.13 10:12