Google plans Gemini 3 launch in 2025 as AI drives USD 100B revenue
Google is gearing up for Gemini 3, its next flagship model that’s penciled in for a 2025 launch. The rollout fits a pattern we’ve seen before, each Gemini version tends to drop near the end of the year, and analysts have been watching that calendar like a clock. Sundar Pichai, though, keeps the hype in check, reminding investors that progress with huge models usually comes in small steps, not overnight miracles.
Around the same time, Alphabet’s numbers got a serious boost. In Q3 2025 the company posted revenue over $100 billion for the first time, roughly double what it earned five years ago. Most of that jump seems to stem from the growing appetite for generative-AI services, with Gemini at the core of the plan.
As we wait for the new model, the pressure is obvious: the next release has to live up to the sky-high hopes that followed an era of record earnings.
Historically, Google has released new Gemini versions in December. Generative AI fuels record growth at Google Gemini is now central to Google's AI ambitions. Alphabet's quarterly revenue topped $100 billion for the first time in Q3 2025, doubling in five years.
"We're firmly in the generative AI era," Pichai told investors. The Gemini app now counts over 650 million monthly active users, with queries tripling since the last quarter. Still, Gemini trails ChatGPT, which OpenAI reports has 800 million weekly users.
Google Search is seeing major gains from AI features, especially with younger audiences. AI Overviews and AI Mode have driven up search volume, and AI Mode--now available in 40 languages--has hit 75 million daily users.
Google says Gemini 3 is slated for a 2025 release, keeping its habit of dropping new Gemini versions in December. Sundar Pichai made it clear the model isn’t expected to deliver a sudden leap; he hinted we’ll see capabilities added bit by bit. The goal, apparently, is to edge out Gemini 2.5 Pro and close the gap with OpenAI’s GPT-5 on tasks that blend agents and multimodal input.
Alphabet’s latest numbers show generative AI now supports about $100 billion of quarterly revenue - roughly double what it was five years ago. Still, it’s hard to say how much of that will stick as market share for Gemini 3. Investors will probably wait for real-world tests before forming a verdict.
For me, the news feels like a vote of confidence in the AI push, yet the model’s true performance and commercial pull are yet to be proven. The roadmap does mention more tweaks after launch, but Google hasn’t given a timeline for those updates.
Common Questions Answered
When is Google planning to launch Gemini 3 and how does this fit with its historical release pattern?
Google aims to launch Gemini 3 in December 2025, continuing its tradition of unveiling new Gemini versions toward the end of the year. This cadence has been observed for previous Gemini iterations and aligns with analysts' expectations for a year‑end rollout.
What financial milestone did Alphabet achieve in Q3 2025 and how is generative AI linked to this growth?
Alphabet reported quarterly revenue exceeding $100 billion in Q3 2025, marking the first time it crossed that threshold and representing a doubling of revenue over the past five years. Sundar Pichai attributed this surge largely to the generative AI boom, with Gemini playing a central role in driving the growth.
How does the user engagement of the Gemini app compare to its previous performance?
The Gemini app now boasts over 650 million monthly active users, and the number of queries has tripled since the prior quarter. This rapid increase underscores the app’s expanding adoption and its importance within Google’s AI ecosystem.
In what ways does Gemini 3 aim to close the performance gap with OpenAI’s GPT‑5?
Gemini 3 is designed to surpass Gemini 2.5 Pro and narrow the gap with GPT‑5, particularly in agent‑like, multimodal tasks such as handling diverse data types and complex interactions. While Sundar Pichai cautions that breakthroughs will be incremental, the new model targets stronger performance in these advanced use cases.