Google I/O 2025: Key Announcements and Highlights

Google I/O 2025: Key Announcements and Highlights

Google I/O 2025, the company’s premier event for developers, featured major updates and product announcements across Google’s ecosystem.

Here are the most significant highlights from the conference:

• A year ago, Google was processing 9.7 trillion tokens per month across its products and APIs. Today, that number has skyrocketed to over 480 trillion — a 50-fold increase.
• More than 7 million developers are now building with Gemini, five times more than a year ago. Gemini usage on Vertex AI has surged 40-fold.
• The Gemini app now boasts over 400 million monthly active users. Engagement is particularly strong with the Gemini 2.5 series. Among users of 2.5 Pro within the app, usage has increased by 45%.
• Project Starline becomes Project Beam: Beam uses a cutting-edge video model to transform 2D video streams into a realistic 3D experience. It employs six cameras and AI to merge video feeds and render participants on a 3D lightfield display.
• Google Meet will introduce real-time live voice translation. English ↔ Spanish translation will begin testing for Workspace users this year, with more languages to follow.
• The Gemini API and SDK are now compatible with MCP tools.
• Project Manier is bringing multitasking AI agents to personal computers through new 'tech and repeat' technology.
• A new Google Search AI Mode is rolling out in the U.S. today. This mode uses Google's "query fan-out" technique to break your question into subtopics and launch multiple simultaneous queries. The result is deeper, more relevant web searches that surface high-quality, hyper-relevant content.
• Gemini 2.5 Flash now includes a new feature called Deep Think, leveraging Google's latest research in reasoning and parallel thinking techniques.
• For creators, Google is launching FLOW, alongside Veo and Imagen. FLOW allows users to generate cinematic clips and expand short videos into longer, coherent scenes.

It's clear that Google is solidifying its position in AI and setting new standards — not only in the much-discussed AI video space with Veo 3, but across hardware, software, and AI at large.

‍