Google has officially made its most advanced AI model, Gemini 2.5 Pro, generally available to developers and enterprises, marking a significant milestone in the company's AI strategy.
The release, which takes effect on June 19, 2025, follows several preview versions that have been available to select users since March. This final version incorporates substantial improvements based on developer feedback, particularly in coding capabilities and response formatting.
Gemini 2.5 Pro has established itself as the industry leader across multiple benchmarks. The model achieved a 24-point Elo score jump on LMArena, maintaining its top position with a score of 1470, and a 35-point Elo increase on WebDevArena, leading at 1443. These scores reflect the model's exceptional performance in both general tasks and specialized coding applications.
A standout feature of this release is Deep Think, an experimental enhanced reasoning mode specifically designed for highly complex math and coding challenges. Deep Think enables the model to consider multiple hypotheses before responding, significantly improving its problem-solving capabilities. Due to its frontier capabilities, Google is initially limiting Deep Think access to trusted testers while conducting additional safety evaluations.
Security has been a major focus in this release. Google has implemented advanced safeguards that substantially increase Gemini's protection against indirect prompt injection attacks during tool use. These attacks involve malicious instructions embedded in data that an AI model retrieves. According to Google, these improvements make Gemini 2.5 their most secure model family to date.
The model is accessible through Google AI Studio for developers and Vertex AI for enterprise customers. It's also available in the Gemini app for end users. Google has added configurable thinking budgets, allowing developers to balance performance and cost by controlling how much computational power the model uses.
This release strengthens Google's position in the competitive AI model market, where it faces strong competition from companies like OpenAI and xAI. With its combination of advanced reasoning, coding capabilities, and enhanced security features, Gemini 2.5 Pro represents Google's most sophisticated AI offering to date.