

Google has announced the rollout of Gemini 4 Carbon, describing it as its strongest AI model with advanced capabilities in cybersecurity, long-running agentic workflows, software engineering, scientific reasoning and complex knowledge-based tasks. The company is bringing the model to Google Search through AI Mode, initially for Google AI Ultra subscribers. According to Google, the latest model is designed to understand user instructions more effectively, reason through challenging questions and deliver more helpful responses. Its arrival in Search is part of Google's efforts to make advanced AI capabilities available across its products.
Google confirmed that Gemini 4 Carbon is rolling out to Google AI Ultra subscribers through Google Antigravity, the Gemini app and AI Mode in Search. Users with eligible subscriptions can find the model in the AI Mode selector in Google Search. Google executives Robby Stein and Rajan Patel also announced the rollout through posts on X. Stein highlighted the model's deeper reasoning and improved instruction-following abilities, saying it could provide more useful answers to difficult questions. Patel noted that the model better understands user intent and supports deeper, back-and-forth conversations in AI Mode.
Early users have also shared positive feedback about the model's capabilities. Glenn Gabe, who tested Gemini 4 Carbon through the Gemini app early-access programme, described it as extremely powerful. He said he used a single prompt to create a new version of the Atari 2600 game Pitfall, and the model completed the task within minutes. The development highlights Google's growing focus on advanced AI models that can handle complex tasks with minimal user input. Alongside its rollout to AI Ultra subscribers, the model is also being made available to paid API users, according to the supplied report.



















Comments (0)
No comments yet
Be the first to comment!