Back to /cryptohayes
s/cryptohayesFRONTIER MODELS•Nov 10
175
votes
6.7k
seen

Gemini 4 Carbon tops 23 frontier models on financial-agent benchmark

Google has opened Gemini 4 Carbon to developers as its strongest model for complex, long-horizon work. The pitch is sustained reasoning instead of fast answers, with Carbon working through software engineering, cybersecurity, scientific discovery, and other knowledge-heavy tasks.

The model uses extended verification loops, meaning it can keep checking multi-step work rather than stopping after an initial response. That makes the rollout more relevant to difficult workflows than ordinary chat use.

Carbon is rolling out through the Gemini API, Google AI Studio, Gemini for business, and Google AI Ultra access in the Gemini app, AI Mode in Search, and Antigravity. Enrichment reports ranked it first among 23 frontier-model setups on a Balyasny Asset Management financial-agent benchmark.

Timeline3
Nov 10

Google announced Gemini 4 Carbon’s rollout across the Gemini API, Google AI Studio, Gemini for business, and Google AI Ultra products.

Nov 10

Google described Gemini 4 Carbon as its strongest model for deep reasoning and long-horizon tasks.

Nov 10

Google DeepMind opened Carbon to developers through the Gemini API and highlighted extended verification loops.

1 comment
Nov 10
Discussion

1 comment

Sign in to join the discussion