Google unveiled an updated preview version of Gemini 2.5 Pro on 5th June, which achieved a 24-point Elo rating increase on the LMArena leaderboard, maintaining its leading position with 1470 points. On the WebDevArena leaderboard, it rose to 1443 points with a 35-point Elo increase, securing the top position.
The models in the Gemini 2.5 family are reasoning models capable of thinking through their thoughts before responding, resulting in enhanced performance and improved accuracy. According to Tulsee Doshi, Director of Product Management, the latest 2.5 Pro model continues to deliver excellent performance in coding, leading challenging coding benchmarks such as Aider Polyglot. The model demonstrates peak performance on GPQA and Humanity's Last Exam benchmarks, which evaluate models' mathematical, scientific, knowledge, and reasoning capabilities.
Google introduced the Gemini 2.5 Flash-Lite preview version on 17th June, which offers the lowest cost in the 2.5 model family, whilst updating Gemini 2.5 Flash pricing to $0.30 per million input tokens and $2.50 per million output tokens. According to CEO Sundar Pichai's post on the X platform, the model shows better performance in coding, reasoning, science, and mathematics, as well as providing improved style and response structure.
Sources:
1.
Tulsee Doshi
2.
Shrestha Basu Mallick
3.
Our latest Gemini 2.5 Pro update is now in preview.
— Sundar Pichai (@sundarpichai) June 5, 2025
It’s better at coding, reasoning, science + math, shows improved performance across key benchmarks (AIDER Polyglot, GPQA, HLE to name a few), and leads @lmarena_ai with a 24pt Elo score jump since the previous version.
We also… pic.twitter.com/SVjdQ2k1tJ
