Google Releases Gemini 4 Carbon, Claiming Significant Improvements in Coding and Agentic Tasks
Two weeks after the release of Argon, Google announced its next-generation frontier model Gemini 4 Carbon, claiming significant improvements in coding, agentic tasks, and cybersecurity defense. Carbon ranks first with 59 points on the updated Artificial Analysis Intelligence Index v5.0, and rose 89 elo to 1769 points on WebDev Arena, ranking fourth. On safety, Google says Carbon reduces unsupported claims by about 2x compared with Argon, reduces excessive overreach behavior by 26%, and reduces under-service by 29%. The model will first be made available to cybersecurity defenders this week, followed by rollout to developers, enterprises, and consumers.
Why read this
SelectedAI score 78
The original text lists multiple benchmarks and internal test changes for Carbon in coding, agentic tasks, and safety alignment, which can be used to judge the focus of this generation of models.
Topics
Body
Following the launch of Argon two weeks ago, Google today announced Gemini 4 Carbon as its next frontier model.
There are “significant gains” in coding, agentic tasks, and cybersecurity defense (finding vulnerabilities). Notably, “Carbon ranks #1 on the recently updated v5.0 of Artificial Analysis Intelligence Index at 59.”
Related: Google adding more Connected Apps to Gemini & Search with new dev platform
Google says this fast release is due to its researchers and models “working hand in hand to advance capabilities faster.”
“With improved quality of output and artifacts in web development, Carbon jumps +89 elo points to land at #4 (1769) on WebDev Arena.”
For cybersecurity:
“firm lead in SEC-bench Pro (88.2%) for finding vulnerabilities, ExploitGym (54.8%) for exploitation, and CWE-bench (71.0%) for patching.”
“On internal benchmarks, Carbon drives a strong uplift in generating secure code, and in live coding environments with our own teams it shows reduced misalignment and fabrication rates.”
On the safety front, Google worked to improve alignment and trained Carbon to “respond to complex or high-pressure tasks in the right way.”
Greater transparency and honesty: In our internal testing, Carbon cuts unsupported claims by approximately 2x over Argon. It puts methodological caveats directly into the deliverable or withholds blocked actions with a clear list of what remains required, rather than fabricating or hedging around missing evidence.
Improved collaboration with the user in the face of ambiguity: Rather than guessing or forcing a path forward, Carbon retrieves the relevant sources and surfaces discrepancies to the user before acting — as a result, users should observe a much better collaborative experience with Carbon.
Following through without overstepping: Carbon simultaneously reduced overreach by 26% and under-service by 29% resulting in much more reliable execution of multi-step plans end-to-end while respecting project constraints and staying strictly within scope.
Officially, Gemini 4 Carbon is “coming soon to developers, businesses, and consumers.” The model will first be available to “cyber defenders” this week.
Updating…