Has Gemini surpassed ChatGPT? We put the AI models to the test.

Ars Technica conducted a comparative test between the default, non-subscription models of Google Gemini (3.2 Fast) and OpenAI ChatGPT (5.2), prompted by Apple's decision to integrate Gemini into Siri. The evaluation used complex prompts covering dad jokes, math problems, creative writing, biography generation, drafting difficult emails, medical advice, video game guidance, and emergency plane landing instructions. Gemini won on four prompts (math, emails, biography, video game guidance), while ChatGPT won three (dad jokes, creative writing, landing a plane), with medical advice being a tie. Although Gemini scored more points, ChatGPT was deemed more practical in the emergency landing scenario by prioritizing safety over direct instruction. However, Gemini generally showed fewer factual errors and better clarity on informational tasks, suggesting Google has significantly closed the gap with OpenAI since previous tests.

(Source：Ars Technica)