Google Expands Gemini 4 Carbon AI Access For Developers, Businesses With Advanced Coding, Cybersecurity And Research Capabilities

New Delhi: Google has expanded access to its Gemini 4 Carbon artificial intelligence (AI) model to developers, businesses, scientists and consumers, as it seeks to bring advanced AI capabilities to tasks such as software development, financial research and cybersecurity that require multiple steps and extended reasoning.

In a blog post on Tuesday, Google DeepMind described Carbon as its strongest model for cybersecurity and long-running AI workflows across knowledge work, software engineering and scientific research.

"Carbon is our strongest model with frontier performance in cybersecurity capabilities and long-horizon agentic workflows across complex knowledge work, software engineering, and scientific reasoning," the post said.

Carbon is designed for tasks that require more extensive reasoning and repeated verification, including complex coding, debugging and scientific research. Google said the model can also work alongside other AI agents, allowing different systems to handle separate parts of a larger task.

The model is available through the Gemini API in Google AI Studio and Gemini for business for paying developers and enterprises. Its introductory price is the same as that of Gemini 4 Argon, at USD 2 per million input tokens and USD 10 per million output tokens.

Carbon is intended for more demanding tasks than Argon, which Google said remains suitable for most software engineering and knowledge work. Carbon is designed for complex coding, debugging and open-ended research that require deeper reasoning and repeated verification.

Google said Carbon ranked first on CursorBench 4.0, Perplexity's WANDR and Zapier's AutomationBench, at more than twice the lower cost per task compared with the next-best external model.

For developers working on complex projects, Google is also previewing a feature called '/swarm' in Google Antigravity. It allows multiple AI agents to work on different parts of a task simultaneously, with Carbon handling roles such as system design, coordination and auditing.

The company also highlighted Carbon's performance in financial research. In pre-release evaluations, it matched the best model tested on Hebbia's financial research benchmark and ranked at the top among 23 model configurations on Balyasny Asset Management's financial intelligence benchmark.

In cybersecurity testing, Google's Project Fortify system achieved 100 per cent recall in vulnerability scanning and 95 per cent precision in triage, according to the company.

The wider rollout follows the model's earlier limited release to selected cybersecurity defenders through Google's Fairwind programme. Google said access for AI Ultra subscribers through Google Antigravity, the Gemini app and AI Mode in Search is also being rolled out. 

Level up your brain the fun way. Play daily games and quizzes on FPJ Smart!

(Except for the headline, this article has not been edited by FPJ's editorial team and is auto-generated from an agency feed.)