Gemini Spark: Powerful Agent, Narrow Gate
Gemini Spark is Google’s always-on, agentic AI assistant that runs in the background, connects to your Workspace data, and carries out multi-step tasks on your computer and across devices without constant user prompts, acting more like a digital coworker than a simple chatbot.
Google has upgraded Gemini Spark with computer-use skills and integration with the Gemini app for Mac, so it can complete tasks on your behalf. After a limited preview, the company is now broadening Gemini Spark access to more markets and more paying subscribers. In the US, it is rolling out to users on the Google AI Pro subscription, which costs USD 20 (approx. RM92) a month. Globally, access is tied to the AI Ultra subscription, which starts from USD 100 (approx. RM460) per month and can go up to USD 200 (approx. RM920) depending on the plan, bringing higher rate limits, extra storage and other perks alongside Spark. The headline is clear: Gemini Spark is expanding, but only for those willing to pay a premium.
Pricing Sends a Message: Agentic AI Is a Luxury Feature
Gemini Spark pricing is not an afterthought; it is the strategy. Google is making its agentic AI assistant available to more people, but still excludes free users entirely. In the US, Spark now comes with the USD 20 (approx. RM92) Google AI Pro tier, while everyone else around the world must pay for AI Ultra, starting at USD 100 (approx. RM460) and ranging up to USD 200 (approx. RM920) per month depending on the market and plan. That structure positions agentic AI access squarely as a premium benefit, unlike basic AI chatbots that are widely available for free.
“Google’s AI Ultra plan starts from USD 100 (approx. RM460) per month and provides higher rate limits for Gemini models, additional storage, and other perks.” Tying Gemini Spark to this top tier turns the agent into a justification for high subscription prices, not a mainstream capability. The result is a deliberate split: casual users get conventional assistants; paying professionals and businesses get a digital worker that can operate across their apps and files. This may help Google defend margins, but it slows the democratization of advanced agents.
What Spark Can Do—and Who Gets Left Out
On paper, Gemini Spark looks like the kind of tool that could change daily work for a wide audience. The agent is based on Gemini 3.5 and is designed to run in the background, monitoring information and acting across devices. Connect it to a Workspace account and it can execute complex, multi-step actions such as finding travel deals and booking trips, or creating reusable skills that learn your writing style and apply it to future emails. Users can ask Spark to check email and calendar schedules each morning, prioritize tasks, auto-draft replies when certain senders email, summarize long threads, or generate detailed reports in Docs from meeting notes and emails.
These features are exactly the kind of automation that could save time for students, freelancers, and small teams, not just enterprises. Yet the agent remains out of reach for free users and many regions. Spark is rolling out to AI Ultra subscribers globally with local language support, but people in the European Economic Area, the UK, Switzerland, and Nigeria are explicitly excluded from this expansion for now. In effect, the users who might benefit the most from automation are told to either wait—or pay up.
Expansion, Competition, and the Risk of Over-Pricing Agents
Google AI expansion around Gemini Spark shows the company is betting that always-on agents will become central to how people work. Spark is now rolling out to AI Pro subscribers in the US and to AI Ultra users globally, and Google’s support pages say it is “rapidly expanding access to more users and businesses over the coming weeks.” There is also an expectation that, after this international push, AI Pro subscribers outside the US may get Spark within weeks. The direction of travel is obvious: agents everywhere—but not for everyone.
At the same time, the premium framing looks shaky against rivals. Competing agents such as ChatGPT Work and Claude Cowork are also paywalled and start around USD 20 (approx. RM92) per month, but they are already available in some regions that currently cannot access Gemini Spark. By locking its agent behind AI Ultra in most of the world, Google risks pushing price-sensitive users—and even some businesses—toward alternatives that offer agent features at the lower tier Google reserves for the US only. Over-pricing agentic AI access could slow adoption right when the market is forming habits and loyalties.
Conclusion: A Strong Agent, a Narrow Funnel
Gemini Spark looks like a credible, practical agent that can monitor your digital life, summarize it, and act on your behalf across devices and Workspace. It offers daily task summaries based on your calendar, documents, notes and inbox, and reusable skills that should become more helpful over time. The problem is not capability—it is who gets to benefit from it.
By tying Gemini Spark pricing to USD 20 (approx. RM92) subscriptions in one market and to USD 100–200 (approx. RM460–RM920) AI Ultra plans almost everywhere else, Google has made a clear statement: advanced agents are a premium upsell, not a default part of the AI experience. That may help short-term revenue, but it risks ceding the narrative—and the user base—to competitors that treat agentic AI as a standard feature of their entry-level paid plans. If agents are the next interface, Google is building a powerful one, but it is asking many users to stand outside and look in.
