Google has announced Gemini 4 Carbon, which it calls its strongest model, with state-of-the-art cybersecurity and long-horizon capabilities. The announcement came on October 28, 2026, from Koray Kavukcuoglu, SVP at Google DeepMind and Google’s Chief AI Architect.
Ranked number one on a key AI index
Google says Carbon ranks #1 on the recently updated v5.0 of the Artificial Analysis Intelligence Index, with a score of 59. It also reports big gains in coding and agentic tasks, where AI works through multi-step jobs on its own. On WebDev Arena, Carbon gained 89 Elo points to land at #4 with 1768.
Strong results in cyber defense
Google highlights three security benchmarks, all measured without guardrails:
- SEC-bench Pro (finding vulnerabilities): 88.2%
- ExploitGym (checking whether a flaw can be exploited): 54.8%
- CWE-bench (patching the root cause): 71.0%
Google also says Carbon delivers more than double the success rate of Gemini 4 Argon at generating secure code on internal benchmarks. It says the Chrome team found Carbon performed more than 10 times better than another lab’s large frontier model at producing high-quality patches. Cloud security platform Wiz reports Carbon leading its web penetration testing benchmarks.
Built for days-long engineering work
Google says its teams have seen Carbon run for days on hard engineering problems, such as tracking down and patching subtle bugs across large codebases and low-level systems. It describes the model as built for sustained, long-horizon agentic workflows.
Safety claims
Google says Carbon builds on the safety stack used for Argon and reports these internal results:
- Successful cheating on Terminal-Bench 4.0 reduced to 0.0%, with cheating attempts at 0.4%
- High-severity misalignment alerts in live coding environments down by almost 80%
- Unsupported claims cut by roughly half compared with Argon
- Overreach down 26% and under-service down 29%
- A 0.0% attack success rate on Gray Swan’s Indirect Prompt Injection benchmark, compared with 0.7% for Argon
Limited rollout for now
Access is deliberately narrow. Starting this week, Google is making Carbon available to a subset of cyber defenders. It will gradually add trusted partners through its Fairwind Program. Google says it is taking part in the U.S. government’s voluntary process for pre-release model testing, and is still gathering feedback and refining guardrails before opening Carbon to developers, businesses and consumers “as soon as possible.”
Pricing
Carbon launches at an introductory price of $2 per million input tokens and $10 per million output tokens, the same as Argon. Cached input tokens are priced at 95% off the input price. After the introductory period, the price rises to $4 per million input tokens and $20 per million output tokens.
The pace of releases
Carbon arrives about two weeks after Gemini 4 Argon. Google says its researchers and models are working together to move faster while strengthening safeguards at the same time.
Source: Google’s announcement, “Gemini 4 Carbon: our next frontier model,” October 28, 2026. All benchmark figures are Google’s own and have not been independently verified.
