gemini-4IMAGE
Argon is the only model above 50% false claims on Theo's codebase analysis bench
Theo says Gemini 4 Argon performed poorly on his codebase analysis benchmark, making it the only model with a false claim rate above 50%.
Theo - t3.gg
@theoArchived 14 days agoOriginally posted 2026.10.14
Theo - t3.gg @theoOriginally posted 2026.10.14
Just ran Argon on my "codebase analysis" bench. Not looking good. Only model with a >50% false claim rate 🙃
This post quotes another post: “Theo tries Gemini 4 Argon: perfect Skatebench score but trust issues remain”
x.com/theo/status/2110109794355032231Archived 2026.10.14 10:21