Advertise
AI

Google's Gemini 4 Carbon Reportedly in External Testing

Early testers say the model outperforms Gemini 4 Argon on 3D tasks but runs two to four times slower and remains weak at UI generation.

Google's Gemini 4 Carbon model has reportedly entered external partner testing, including trials on the LM Arena model comparison platform, according to aitimes.com. Early testers describe better 3D output than the earlier Gemini 4 Argon model, but slower performance.

Testers say Carbon takes two to four times as long as Argon on some tasks. They also report that its webpage design and UI generation remain weak.

Google has not officially confirmed Carbon's existence, performance or release schedule. The reported testing activity follows earlier signs of internal use: in October 2026, Business Insider reported that Google employees were testing the model in the company's internal coding tool, JetSki.

Gemini 4 Argon was released about a week before Carbon outputs began to appear.

Quick answers

Has Google confirmed Gemini 4 Carbon?

No. Google has not officially confirmed the model's existence, performance or release schedule.

How does Gemini 4 Carbon compare with Gemini 4 Argon?

Early testers say Carbon handles 3D tasks better but is slower, taking two to four times as long on some tasks, and that its webpage design and UI generation remain weak.

Source