Google expands access
Google introduced Gemini 4 Carbon on October 28, 2026, initially for selected cybersecurity defenders. On November 10, reporting indicates wider access for paying API developers and eligible Google AI Ultra subscribers. Availability may vary by product, region and account.
Built for difficult engineering work
Google positions Carbon for long-running engineering, debugging, security and multi-step reasoning. The company says the model improves on coding and cyberdefense benchmarks. These are reported benchmark results, not a guarantee for every application.
Pricing and practical considerations
Google announced an introductory API price of $2 per million input tokens and $10 per million output tokens, with higher prices planned after the introductory period. Teams should confirm live prices and access in Google AI Studio before deployment.
What to watch
Developers should compare Carbon with existing models using real codebases, monitor token costs, and evaluate security, latency and reliability before switching production workflows.
Sources
Google official announcement: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/ . Coverage of November 10 expanded availability: https://www.unite.ai/google-expands-gemini-4-carbon-access-beyond-cybersecurity-defenders/ .