Summary
Google has unveiled Gemini 4 Carbon, a new frontier model that builds directly on Gemini 4 Argon, released just two weeks earlier. This video breaks down what Carbon is, what Google claims it can do, and who can use it right now.
Google highlights two areas of strength: cybersecurity and long-horizon work. Internal teams reportedly saw the model run for days on complex engineering tasks without giving up on dead ends. On the security side, Carbon is built to find vulnerabilities, confirm them, and patch them, and it leads all three of Google’s security tests. According to the video, it beats GPT-6 Astra, Claude Opus 5.5, and Claude Fable 5.1 on 16 of the 18 benchmarks Google shared, trailing only on a coding test and a computer-use test. It also posted a 0% attack success rate on Gray Swan’s prompt injection benchmark and reportedly makes about half as many unsupported claims as Argon.
Carbon is not yet generally available. A small group of cyber defenders gets access first, followed by trusted partners, with no public date announced. The video points to Argon’s timeline, which went from cyber defenders on September 30 to public release on October 13, as a possible guide. It also touches on confusion over Google’s Argon and Carbon naming and how the release pace compares with OpenAI and Anthropic.
📺 Source: Paul J Lipsky · Published October 28, 2026
🏷️ Format: News Analysis
