Gemini 4 Carbon Frontier AI Model Falls Slightly Behind Claude Opus 5.5 in Agentic Capabilities, but Not in Cost Efficiency

Google’s Gemini 4 Carbon was the second best at performing agentic tasks on SemiAnalysis’ recently released GDPx benchmark test. The frontier AI model managed to score 10.9 percentage points, which was slightly lower than the 12-percentage-point score of Anthropic’s Claude Opus 5.5. While Google’s AI model came in second position, it was significantly more cost-efficient than the Claude Opus 5.5 AI model. With 5,94,000 total output tokens, the Gemini 4 Carbon frontier AI model completed each task at $16.29. Meanwhile, Anthropic’s AI model cost $51.41 per task with a total of 5,48,000 output tokens.

More in Gadgets

Shooting Video With Honor’s New Magic 9 Pro Max Lets You Feel Like a Movie Director November 4, 2026 iPhone 18 Key Specifications Leaked by Tipster Ahead of 2027 Launch; May Get a Price Hike November 4, 2026 Google Is Forcing Schools to Choose Between Its Search Engine and AI November 4, 2026 AI Data Centers Have a PR Problem. OpenAI Is Proposing a Rebrand November 4, 2026 Roku Retires Aquatic Life Screensaver to Reel Viewers Into Roku City November 4, 2026 Apple Is Putting an Age Limit on Falling in Love With AI November 3, 2026

See all 198 in Gadgets →