Google Gemini 2.5 Is the Newest Model Set To Compete With DeepSeek R1

Imad is a senior reporter covering Google and internet culture. Hailing from Texas, Imad started his journalism career in 2013 and has amassed bylines with The New York Times, The Washington Post, ESPN, Tom's Guide and Wired, among others.

Gemini 2.5 is a new AI reasoning model that's set to compete with DeepSeek R1 and is currently the highest-rated AI model overall on LMArena, Google said in a blog post on Tuesday. 

Google describes the new line of Gemini 2.5 models as "thinking models," ones that recursively analyzes their answers before giving users a final output. Per benchmarks on LMArena, Gemini 2.5 is leading in reasoning, science, math and agentic coding. It's not winning in all tests, however. For example, OpenAI o3-mini still leads it in LiveCodeBench v5. 

Gemini 2.5 is rolling out to paid Advanced ...