Google I/O 2025: AI Takes Center Stage with Gemini 2.5 and New Search Features

Google's annual developer conference delivered a broad portfolio of AI-powered innovations this past week, reshaping how we'll interact with Google search, productivity tools, and creative applications. Here's everything you need to know about the biggest announcements from Google I/O 2025.

TL;DR

Google is accelerating AI integration across all its products, with Gemini 2.5 now powering enhanced search experiences, new subscription tiers offering premium AI access, and revolutionary tools for coding, creativity, and productivity. The company is clearly positioning itself as the leader in practical AI applications.

Gemini 2.5

Gemini 2.5 Pro: Enhanced Intelligence

Gemini 2.5 Pro leads in math and science benchmarks like GPQA and AIME 2025, scoring a state-of-the-art 18.8% on Humanity's Last Exam. The model has achieved significant improvements in coding performance, with 63.8% on SWE-Bench Verified, the industry standard for agentic code evaluations.

Key capabilities include:

• Deep Think Mode: An "enhanced reasoning" mode that leads across benchmarks for Mathematics, Code, and Multimodality

• LearnLM Integration: Gemini 2.5 Pro is now the world's leading model for learning, outperforming competitors on every category of learning science principles

• Native Audio Output: More natural conversational experiences with customizable tone and accent

Gemini 2.5 Flash: Speed Meets Quality

The new 2.5 Flash is improved across key benchmarks for reasoning, multimodality, code and long context while getting even more efficient, using 20-30% less tokens. It's second only to 2.5 Pro on the LMArena leaderboard and becomes the new default model starting in June.

Search: AI Mode Goes Mainstream

AI Mode Expansion

AI Mode is now rolling out to all US users, expanding what AI Overviews can do with more advanced reasoning, thinking and multimodal capabilities. This is Google's direct answer to ChatGPT's search capabilities.

What makes AI Mode special:

• Custom Gemini 2.5: Uses a custom version of Gemini 2.5 with a multistep approach to make a plan, conduct searches to find information and adjust the plan based on what it finds

• Real-time Integration: Access to Knowledge Graph, shopping data for billions of products, and fresh information sources

• Follow-up Conversations: Ask complex, multi-part questions and dig deeper with follow-ups

AI Overviews Scale Globally

AI Overviews have scaled up to 1.5 billion monthly users in 200 countries and territories, with Google Search bringing generative AI to more people than any other product in the world. In major markets like the US and India, AI Overviews is driving over 10% increase in usage for the types of queries that show AI Overviews.

New Subscription Tiers

Google AI Ultra

Google AI Ultra costs $249.99 a month, with a special offer for first-time users of 50% off for your first three months. This tier offers:

• Exclusive Access: Early access to cutting-edge models like Veo 3 and Deep Think mode

• Agent Mode: A new experimental capability where you simply state your objective, and Gemini intelligently orchestrates the steps to achieve it

• Project Mariner: Agentic capabilities for automating web tasks like booking tickets and making reservations

• Maximum Limits: Highest usage limits and 30TB of storage plus YouTube Premium

Google AI Pro

Google AI Pro costs $19.99/month and includes products like Flow, NotebookLM and more, all with special features and higher rate limits. Starting tomorrow, Gemini in Chrome will be able to summarize information from any webpage for Google AI Pro subscribers.

Free Access Expansion

College students in the U.S., Brazil, Indonesia, Japan and the U.K. are eligible to get a free upgrade of Gemini for a whole school year, with more countries coming soon.

Creative Tools: Flow and Enhanced Media Generation

Google Flow: AI Filmmaking Revolution

Google Flow is an "AI filmmaking tool" built on Veo, Imagen, and Gemini models, described as an evolution of VideoFX with natural-language prompting, camera controls and a scene builder.

Imagen 4 and Veo 3

• Imagen 4: Better detail and text outputs rolling out to the Gemini app

• Veo 3: Features native audio generation to accompany video, available with Google AI Ultra in the US

Developer and Enterprise Tools

Jules: The Coding Agent

Jules is described as a "true coding agent" that "reads your code, understands your intent, and gets to work" in the background, capable of writing tests, building new features, fixing bugs, and bumping dependency versions. It's available in public beta worldwide with no waitlist.

Project Mariner Integration

Project Mariner's computer use capabilities are coming to developers via the Gemini API, with trusted testers like Automation Anywhere and UiPath already building with it.

Enhanced Development Experience

• Android Studio: The built-in Gemini integration is now powered by 2.5 Pro

• Live API: Introducing audio-visual input and native audio out dialogue for building conversational experiences

Productivity and Communication Enhancements

Gemini Live Expansion

Gemini Live with camera and screen sharing is now available to everyone on Android and iOS for free, with apps/Extensions support coming to Gemini Live starting with Calendar, Keep, Maps, and Tasks.

Google Meet Translation

Gemini will power a new speech translation feature in Google Meet, starting with English and Spanish, rolling out to Google AI Pro and Ultra subscribers.

Enhanced Canvas and Deep Research

• Canvas: Create interactive infographics, quizzes and even podcast-style Audio Overviews in 45 languages

• Deep Research: You can now upload your own PDFs and images to inform Deep Research, which gets combined with public data

Safety and Detection Tools

SynthID Detector

Google announced SynthID Detector, a verification portal that helps to quickly and efficiently identify content that is watermarked with SynthID. Since launch, SynthID has already watermarked over 10 billion pieces of content.

Looking Ahead: The Agent Ecosystem

Google is building toward a comprehensive agent ecosystem with tools like the open Agent2Agent Protocol for agents to communicate with each other, and compatibility with the Model Context Protocol so agents can access other services.

The company's rapid pace of innovation is evident in CEO Sundar Pichai's keynote observation: "In our Gemini era, we're just as likely to ship our most intelligent model on a Tuesday in March, or announce a really cool breakthrough like AlphaEvolve a week before".

The Impact:

Google I/O 2025 shows the rapid transition of AI from experimental features to core product experiences. With the Gemini app now having over 400 million monthly active users and AI capabilities being integrated across search, productivity, and creative tools, Google is positioning itself at the forefront of AI capabilities.

Whether you're a developer looking to build with cutting-edge AI capabilities, a business seeking productivity enhancements, or a consumer wanting smarter digital experiences, Google's I/O 2025 announcements offer something transformative for everyone.