gemini-4POST
Gemini 4 Argon benchmarks look strong but coding still trails OpenAI and Anthropic, poster says
The poster says Gemini 4 Argon looks impressive on benchmark charts, but its real coding performance still falls short of OpenAI and Anthropic, and questions whether Google can catch up.
Drew Wakefield
@DrewWakefield5Archived 18 days agoOriginally posted 2026.10.04
Drew Wakefield @DrewWakefield5Originally posted 2026.10.04
Google waited 6 months for a major release, and Gemini 4 Argon looks amazing on benchmark charts. Then you look at the actual coding performance and it still isn't up to scratch with OpenAI or Anthropic. Wild that a company the size of Google with endless billions and compute is constantly playing catch-up. Are Google ever going to catch up on coding, or is the gap just too big?
x.com/drewwakefield5/status/2106772585388097947Archived 2026.10.12 10:05