Google claims Gemini 2.5 Pro preview beats DeepSeek R1 and Grok 3 Beta in coding performance
1 min read
Summary
Google has upgraded its AI tool Gemini 2.5 Pro for the third time in eight months, improving its performance and pushing it closer to enterprise-ready status.
The new version, called Gemini 2.5 Pro Preview 06-05 Thinking, is available to test via the Google AI Studio and Vertex AI platforms.
According to Google, the latest iteration of the tool outperformed competitors including OpenAI’s o3, o3-mini, and o4-mini, as well as models from Anthropic, xAI and DeepSeek in benchmark tests.
It improved by 24 points in the LMArena measurement and by 35 points in WebDevArena.
Google has not confirmed when a general availability version of Gemini 2.5 Pro will launch, but the company has said the current preview is “ready for enterprise-scale capabilities”.