LOGO

Gemini 2.5 Pro: Google Updates AI Model Before I/O

May 6, 2025
Gemini 2.5 Pro: Google Updates AI Model Before I/O

Google Launches Gemini 2.5 Pro Preview (I/O Edition)

On Tuesday, Google announced the availability of Gemini 2.5 Pro Preview (I/O edition). This represents an updated iteration of its leading Gemini 2.5 Pro AI model.

The company asserts that this new version surpasses performance levels on several commonly utilized benchmarks.

Availability and Pricing

Access to Gemini 2.5 Pro Preview (I/O edition) is provided through the Gemini API, as well as Google’s Vertex AI and AI Studio platforms.

Importantly, the pricing structure remains consistent with the Gemini 2.5 Pro model it is designed to supersede.

The updated model is also integrated into Google’s Gemini chatbot application, accessible via web browsers and mobile devices.

Anticipating Google I/O

This release strategically precedes Google’s annual I/O developer conference. The “I/O edition” designation reflects this timing.

During the conference, Google is anticipated to introduce a range of new models, alongside innovative AI-driven tools and platforms.

The AI landscape is intensely competitive, with companies like OpenAI and xAI poised to release high-performing models.

Enhanced Capabilities

Google reports that Gemini 2.5 Pro Preview (I/O edition) demonstrates “significantly” improved capabilities in the areas of coding and the creation of interactive web applications.

The model exhibits enhanced proficiency in tasks such as code transformation – adapting code to meet specific objectives – and code editing.

Benchmark Performance

According to a recent blog post from Google, Gemini 2.5 Pro Preview (I/O edition) currently leads the WebDev Arena Leaderboard.

This benchmark evaluates a model’s capacity to generate visually appealing and functionally sound web applications.

Furthermore, the model achieves state-of-the-art performance in video understanding, attaining a score of 84.8% on the VideoMME benchmark.

Developer Benefits

Google highlights that this new version will not only enhance coding performance for existing Gemini 2.5 Pro users.

It will also address valuable developer feedback, specifically reducing errors in function calling and improving the rate at which functions are triggered.

The model is designed with a preference for aesthetically pleasing web development, while still allowing for precise control and customization.

#Gemini 2.5 Pro#Google AI#AI model#Google I/O#artificial intelligence