A TechRadar test pitted Gemini 3 against ChatGPT 5.1 and Claude Sonnet 4.5 in a coding challenge involving a simple web game. Gemini 3 emerged as the top performer, delivering a surprising result.
In a recent evaluation published by TechRadar, the author compared three leading AI models: Gemini 3, ChatGPT 5.1, and Claude Sonnet 4.5. The focus was on their ability to handle a real-world coding task—creating a silly web game.
The test revealed that Gemini 3 'crushed it,' outperforming the others in this practical scenario. This outcome highlights potential strengths in Google's AI for coding applications, though the article describes the task as lighthearted.
The comparison was published on November 18, 2025, emphasizing hands-on performance over benchmarks.