The new model is now available to enterprise clients, paying developers, and Google AI Ultra subscribers across multiple services including Search's AI Mode and Google AI Studio. It introduces an Antigravity feature that deploys coordinated agent teams from single user prompts to handle intricate tasks. Introductory token costs are set significantly lower than competing tools from OpenAI and Anthropic, though rates are scheduled to double in late November. Despite these performance gains, benchmark tests place the system just behind Anthropic's recently launched flagship offering.
Key points
Google is leveraging steep introductory token discounts to aggressively undercut rivals l…: Google is leveraging steep introductory token discounts to aggressively undercut rivals like Anthropic's Fable 5.2 and OpenAI's GPT-6 Astra.
Access remains limited to enterprise accounts, paying developers, and top-tier subscriber…: Access remains limited to enterprise accounts, paying developers, and top-tier subscribers, leaving availability for free-tier users currently unconfirmed.
Benchmark evaluations position Carbon among the industry's elite models for complex tasks…: Benchmark evaluations position Carbon among the industry's elite models for complex tasks, though it trails Anthropic's latest flagship release.
Sources checked by our analysis pipeline: 1
Google is rolling out its latest AI model, Gemini 4 Carbon, to the general public after releasing it to a select group of “cybersecurity defenders” two weeks ago. Google claims Carbon is its “strongest model” yet, and capable of competing with the very best systems from Anthropic and OpenAI.
Google releases Gemini 4 Carbon with agent ‘swarm’ feature
It ‘pushes the frontier even further’ than its last release, Argon.
It ‘pushes the frontier even further’ than its last release, Argon.
Google says Gemini 4 Carbon “pushes the frontier even further” than Argon, referring to the company’s other frontier model that launched only last month. The company said Argon will likely remain “the ideal choice for the majority of software engineering and knowledge work,” but that Carbon is “built for tasks that test the limits of a model’s intelligence,” such as complex, multi-step coding and open-ended scientific research.
To highlight Carbon’s utility for long-horizon work involving AI agents, Google said it is also previewing a “swarm” feature on its agentic development platform Antigravity, which will let users turn a single prompt…
Meta's Muse and similar digital assistants have recently been obstructed from scanning platform listings, occasionally forcing traffic to external mirror sites. This pushback aligns with broader terms banning automated checkout bots without human oversight to curb unauthorized data scraping. The shift highlights growing friction between marketplace security and the rise of automated consumer tools.
High-profile venture firms and chipmakers continue pouring capital into artificial intelligence infrastructure, rapidly driving valuations higher for foundational technology providers. This latest injection of funds gives the company substantial resources to expand its operations within the increasingly competitive enterprise AI training sector.
Built upon the recently launched Gemini 4 Argon, this new model is tailored for multi-step reasoning, advanced software engineering, and long-horizon agentic workflows. It aims to empower enterprise specialists, researchers, and developers tackling dense data synthesis across legal, technical, and financial environments. Access is rolling out via API integrations as well as dedicated consumer and business subscription tiers.