one year on
Google I/O 2025: Veo 3 generates video with synchronized audio, AI Mode rolls out in US Search
Google unveils Veo 3 with native audio generation, Imagen 4, a $249.99/month Gemini Ultra tier, and AI Mode in Search, in a keynote that cast the company as firing on every AI front.
MOUNTAIN VIEW — Google today used its annual I/O developer conference to announce a wave of new AI capabilities, including Veo 3, a video generation model that produces synchronized audio—marking the first time the company has demonstrated native sound in AI-generated video.
CEO Sundar Pichai said the company is processing 480 trillion tokens per month across its products and APIs, up from 9.7 trillion a year ago. He also announced that Gemini 2.5 Pro now leads the LMArena leaderboard in all categories, and that the Gemini app has over 400 million monthly active users.
The flagship consumer product is a new $249.99-per-month Gemini Ultra tier. Subscribers get access to Veo 3, Imagen 4, the Flow video editing tool, Gemini 2.5 Pro’s Deep Think reasoning mode, and 30TB of storage. Pichai described the tier as “the highest level of access to Google’s AI-powered apps and services.”
The company also rolled out AI Mode in U.S. Search—a tab that allows users to ask longer, more complex queries with follow-up questions, powered by Gemini 2.5. Pichai called it “a total reimagining of Search.” Early testers submitted queries two to three times the length of traditional searches, he said.
On the infrastructure side, Google announced Ironwood, its seventh-generation TPU, which delivers 10 times the performance of the previous generation and 42.5 exaflops per pod. The chip is designed specifically for inferential and thinking AI workloads.
The announcements follow a pattern of rapid releases that Pichai acknowledged directly: “Normally, you wouldn’t have heard much from us in the weeks leading up to I/O… but in our Gemini era, we’re just as likely to ship our most intelligent model on a Tuesday in March.” The message was clear—Google is done saving its best for a keynote.
The record
Reported that Veo 3 claims to generate sound effects, background noises, and dialogue to accompany videos; noted Deep Think mode may be similar to OpenAI's o1-pro and o3-pro.
view the original post →One year later — open only if you can handle spoilers
Veo 3's audio synchronization proved to be a genuine differentiator, though high price kept adoption narrow. The Gemini Ultra tier saw slower uptake than hoped, while AI Mode in Search gradually became a default for power users, reshaping how Google presents its core product.