· via dev.to (home feed)
Google opens Gemini Tasks agent mode to free users in six countries
Google has opened Gemini Tasks, its background agent mode previously limited to AI Pro and Ultra subscribers, to free Gemini users in six countries, with Duplex calling, Google Pay and Gmail integrations arriving US-first.

Gemini Tasks moves to the free tier
On November 17, Google began rolling out Gemini Tasks, its background agent mode, to free Gemini accounts. Until then the capability required an AI Pro or AI Ultra subscription. The rollout covers six markets at launch: the United States, Japan, South Korea, Canada, Australia and India. These details come from a dev.to writeup citing Android Central's report on a November 17 announcement by Google's Josh Woodward.
India's presence on the first-day list is the detail the dev.to author singles out. Two weeks earlier the feature was paid-only there; afterward, any student with a free account could hand off research work while doing homework.
How the agent works
In the Gemini app on mobile and on the web, a toggle above the prompt box lets users switch between Chat and Task. Where Chat responds conversationally, Task accepts the entire assignment and completes it in the background rather than guiding the user through it interactively. Example jobs mentioned include putting together a trip itinerary, researching a topic for a school project, and producing a recurring morning briefing.
The mode runs on Gemini 4 Hydrogen, per the dev.to post, the same fast model Google made the default in Search's AI Mode on November 12.
Real-world actions arrive US-first
The more consequential changes are the actions the agent can take. According to the report, Gemini Tasks can phone local businesses using Google Duplex, complete purchases through Google Pay, receive text messages over RCS, and process emails forwarded to Gmail. Those capabilities are currently limited to the United States.
The mode can also sign into websites on a user's behalf through Google Password Manager. Google's framing, as relayed in the dev.to post, is that passwords are not exposed outside Google's ecosystem. The author is more guarded about this particular feature, arguing that delegating login credentials to an agent demands considerable trust and is worth observing closely before it is pointed at anything important.
A test case for free agents
The dev.to author, who builds automation agents as a hobby, plans to benchmark Gemini Tasks against their own scripts this week: if a free agent can handle routine research steps, the scripts get shorter. That is the practical question for developers as well — whether a zero-cost, first-party agent reduces the case for hand-rolled automation on simple tasks.
Why it matters
Moving an agent mode from paid tiers to free accounts across six markets at once is a distribution decision, not just a pricing one. It puts autonomous, multi-step task execution in front of anyone with a free Gemini account in those countries, including students for whom a subscription was the barrier two weeks ago. It also sets an expectation: if agentic features become part of the free tier at Google, competitors selling agent access by subscription will face pressure to justify the price.
The integration list is the other half of the story. An agent that can call businesses, spend money through Google Pay, read forwarded email and log into websites is no longer a chat wrapper — it acts in the world with credentials attached. That dramatically expands what free users can do, and it concentrates a great deal of trust in a single company's ecosystem, which makes the rollout worth watching on both convenience and oversight grounds.
As with any early rollout, the specifics here rest on one chain of reporting — a dev.to post citing Android Central's account of Woodward's announcement — so details such as country availability and the timing of the US-first integrations may shift as the feature reaches more users.
- #google-gemini
- #ai-agents
- #gemini-tasks
- #free-tier