Google launches Gemini 4 Carbon AI model with cybersecurity focus

Alphabet's Google unveiled Gemini 4 Carbon, its newest frontier AI model, on Wednesday, intensifying the competition with OpenAI and Anthropic as the race to develop superior coding and cybersecurity systems escalates. This launch follows Google's recent introduction of Gemini 4 Argon two weeks prior, as the company broadens its Gemini line of AI models tailored for developers and enterprise customers.

Google claims that the Carbon model exhibits exceptional performance in cybersecurity tasks such as vulnerability detection, exploitation, and software patching. The company states that Carbon generates secure code at a success rate more than double that of Gemini 4 Argon on internal benchmarks. Additionally, the model is intended to manage intricate, long-term engineering and coding tasks.

Currently, Carbon is exclusively accessible to a select group of cybersecurity professionals through Google's Fairwind Program while the company continues testing its protective measures. The AI model will be priced at $2 per million input tokens and $10 per million output tokens, mirroring the introductory pricing of Gemini 4 Argon.