Google plans to bring its always-on Gemini Agent to everyday users this year

What Google said

Google plans to give regular consumers the core features of its business-focused Gemini Agent before the year ends. As aitimes.com reported, CEO Sundar Pichai made the statement on the company's earnings call on the 28th, US time.

The agent takes a goal from the user, plans the steps and carries them out itself. It runs in Google's cloud around the clock and can handle work that spans several services. Pichai said the rollout will begin with the Gemini app, which has 1 billion users, and that related features will also come to Google Search.

The longer-term idea is to link major products such as Chrome and Android into one agent experience. Google gave no date for that part.

Business first, consumers second

Google revealed the Gemini Agent on the 8th as a unified agent for work. A user sets a goal, and the AI makes a plan and uses the tools it needs to produce a finished result. It connects to Gmail, Google Drive, Docs and Calendar, and it can split a big job into steps. It can also spawn helper agents to share the work.

Because it runs in the cloud, it can keep going after you shut your laptop. Administrators can apply an organization's security rules and access permissions to it.

Pichai said Google built the business version first because corporate customers demand the strictest standards on security, governance and control. The plan is to reuse that foundation for consumers, while the company keeps widening access to the business product.

How this differs from Gemini Spark

Google already released a consumer agent called Gemini Spark in May, and some consumers have it. Spark also manages email, calendars and documents, handles recurring tasks and works in Google's cloud around the clock. The two products overlap a lot in how they work.

The main differences are purpose and support features. Spark is aimed at personal life and work. The enterprise agent is designed for shared company systems, with several agents cooperating, long-running jobs, access to company data and security management.

What is unclear is whether this plan replaces Spark, merges with it or becomes a separate product. Pichai did not address that, according to the report.

The hard part: serving a billion people

The most concrete part of the call was the engineering challenge. Pichai said Google is building a lightweight model that delivers the performance of Gemini 4 Carbon more efficiently. The goal is lower computing cost and faster responses without losing the ability to do the work, so it can serve a very large audience.

He also said the model alone is not enough. A harness, meaning the software wrapper that runs and manages a complex task reliably, and a security system are needed too. Google is using its cloud to build an environment where people can leave work running around the clock.

Pichai added that users are already changing how they use the Gemini app. They are asking for multistep jobs rather than simple chat, and the recently added Task tab shows that demand. He said finishing long tasks with a satisfying result is what brings people back. If the success rate is too low, the service feels broken to the user, so Google keeps fixing problems that show up in real use.

What is still unknown

Google also announced a Connected Apps Platform on the same day. It lets outside developers link their services to Gemini and Search, and it manages those connections in one place. In future, a consumer agent could use it to act across many services.

Google did not say, however, that the platform is directly tied to the new consumer agent.

The company also gave no exact launch date, pricing or list of countries. The 1 billion figure describes the Gemini app's existing user base, the starting point for the rollout. It is not a count of people who will use the agent.