Google makes Gemini 4 Carbon generally available
Google DeepMind has opened Gemini 4 Carbon, its strongest model, to developers, enterprises and consumers at Argon's introductory price, and claims top agentic benchmark scores at less than half the cost per task.
As we reported earlier, Google DeepMind unveiled Gemini 4 Carbon two weeks ago and first gave it to cybersecurity defenders through its Fairwind Program. The company has now announced general availability.
Carbon is available through the API in Google AI Studio and Gemini for business to all paying developers and enterprises. It costs $2 per million input tokens and $10 per million output tokens, the same introductory price as Gemini 4 Argon. It is also rolling out to Google AI Ultra subscribers in Google Antigravity, the Gemini app and AI Mode in Search.
Google describes Argon as the default model for most coding and knowledge work. Carbon is meant for harder multi-step coding, debugging and open-ended research, and it uses more tokens to reason and verify its output. Google says Carbon ranks #1 on CursorBench 4.0, Perplexity's WANDR and Zapier's AutomationBench, at more than 2X lower cost per task than the next-best external model. On the new Multimodal Knowledge Agent benchmark, which Google built with Artificial Analysis, Carbon ranks #2 at more than 6X lower average cost per task than the leader.
Google is also previewing /swarm in Antigravity. The feature turns one prompt into a team of coordinated agents, with Carbon handling design, coordination and auditing while lighter models work in parallel.
On security, Google says Carbon reached 100% scanning recall and 95% triage precision on an internal benchmark for its Project Fortify system. Wiz used Carbon to find zero-day vulnerabilities at two major SaaS providers in under two hours. Google says details of separate bugs found by its Cloud Vulnerability Research team will come after responsible disclosure is complete.
None of these benchmark claims has been independently verified.