Home · Technology · Oct 29 archive

Google Launches Gemini 4 Carbon AI Model with Cybersecurity Focus

Updated 2:14 AM ET · 4:46 PM ETConfirmed

Technology Desk

In Short: Google has launched Gemini 4 Carbon, the latest addition to its Gemini family of artificial intelligence models, with a focus on cybersecurity and advanced coding tasks.

The new model, announced on October 28, 2026, is designed to perform a range of cybersecurity functions, including identifying software vulnerabilities, examining potential exploitation methods, and assisting with the development of patches.

According to Google, Gemini 4 Carbon is intended to handle complex, long-running engineering and coding assignments, with a success rate in generating secure code more than twice that of its predecessor, Gemini 4 Argon, based on internal benchmarks.

Pricing for Gemini 4 Carbon is set at $2 per million input tokens and $10 per million output tokens, matching the introductory pricing for Gemini 4 Argon.

The model is initially being made available to a selected group of cybersecurity defenders through Google’s Fairwind Program, with access expanding gradually.

Google says the launch of Gemini 4 Carbon comes amid growing concerns about how AI is changing the cybersecurity landscape, intensifying competition with other AI developers like OpenAI and Anthropic.

The company emphasized that these results are based on Google’s own testing and should not be interpreted as an independent assessment of performance across all cybersecurity environments.

For now, organizations outside the initial group should treat Carbon as an emerging option rather than an immediately available replacement for their existing tools.

What this adds

Background

What's confirmed

• Pricing aimed at developers and businesses Google has set introductory pricing for Gemini 4 Carbon at $2 per million input tokens and $10 per million output tokens, matching the introductory pricing announced for Gemini 4 Argon.
• Google says Carbon is intended to handle complex, long-running engineering and coding assignments.
• The company also says the model can generate secure code at more than twice the success rate of Gemini 4 Argon on its internal benchmarks.

What's still developing

• Reuters The model is the latest addition to Google’s Gemini family and follows the introduction of Gemini 4 Argon approximately two weeks earlier.
• 26 AI plans and models on one page, from official pricing pages, updated every month.
• Its announcement does not mean that everyone can open the Gemini app and start using it.
• The benchmark page retrieved for this article did not independently corroborate Google’s reported Carbon score.
• The interesting question extends beyond whether Carbon can write impressive code.
• An announcement describes a model and its intended direction.
• The launch comes two ​weeks after Google introduced Gemini 4 Argon, ⁠as the company expands its ​Gemini family of AI models for developers ​and enterprise customers, Reuters reported.
• Google unveiled Gemini 4 Carbon, its newest artificial intelligence model, on Wednesday, highlighting major improvements in coding, autonomous agent capabilities and cybersecurity defense.
• Alphabet-owned Google announced the model on Wednesday, two weeks after introducing Gemini 4 Argon, as it expands its Gemini lineup for developers and enterprise customers, Reuters reported.
• The launch adds to the growing competition among major AI developers to build models capable of performing increasingly sophisticated software engineering and cybersecurity work.
• However, Carbon is entering a controlled rollout.

Sources

• Bold Newslink
• Kingy AIlink
• Latest news from Azerbaijanlink
• Newslink
• WarpBeat — background on Google Launches Gemini 4 Carbon With Enhanced Cybersecurity Capabilities link
• WarpBeat — background on Google Launches Gemini 4 Carbon with Enhanced Cybersecurity and Coding Capabilities link
• Mervin Praison — video link