WHAT NEW IN X11.05 THU

AI collections from X · 23 collections · 69,408 posts

gemini-4IMAGE

Gemini 4 Argon scores 40.7 at high effort, behind Astra 5.1's 45

A user ran benchmarks with the agy CLI: Astra 5.1 max scored 45, Opus 5.5 max 41.7, and Gemini 4 Argon 40.7 at high and 36 at medium, with medium and low results to follow.

Paweł Huryn
@PawelHurynArchived 21 days agoOriginally posted 2026.10.14
Gemini 4 Argon scores 40.7 at high effort, behind Astra 5.1's 45
Paweł Huryn @PawelHurynOriginally posted 2026.10.14
Open original on X

Gemini Argon 4 (medium) + agy CLI is in: Astra 5.1 (max): 45 Fable 5.1 (max): 43 Muse Wonder Full 1.0 (max): 42 Opus 5.5 (max): 41.7 Gemini 4 Argon (high, n=3): 40.7 Gemini 4 Argon (medium, n=2): 36 Opus 5.5 (xhigh) 36 I will drop medium (n=3) and low (n=3) effort levels in this thread in a few hours 🧵

x.com/pawelhuryn/status/2110357427585491369Archived 2026.10.15 11:38