Google opens Connected Apps platform to U.S. developers

Google says more than 50 connected apps are already available in Gemini, and those apps will begin reaching Google Search over the next few weeks. The company has also opened a public beta of its Connected Apps Platform to U.S. developers on a waitlist, without giving a date for wider release.

Google says users can “securely link and interact with popular third-party services,” then use those services to take specific actions. Its examples include connecting Canva to design a flyer and YouTube Music to build a playlist.

The commercial stakes are direct in Google's framing: connected apps could shift consumers away from opening individual apps and toward using services through their preferred AI provider. Google says consumer-facing services such as ChatGPT and Meta’s Muse have pursued the idea with varying approaches. The company also argues that a wide range of connected apps matters if an AI provider is to capture and retain attention.

What exactly is Google opening?

Google says the Connected Apps Platform lets any business or developer submit connected apps for distribution through Gemini, Google Search, or both. It separates that platform debut from the earlier-this-year introduction of connected apps in Gemini and Google Search.

Developers follow four stages in Google's workflow. The process covers integration, submission, distribution choice and post-launch monitoring.

• Integrate the service through a cloud-hosted Model Context Protocol (MCP) server, a protocol for connecting developer services to AI features, or through an Android MCP implementation, the Android version of that connection.
• Use the AppFunctions APIs, the integration interfaces for those connections, then register, test and submit through Google Connected Console, the management console for those steps.
• Choose Search, the Gemini app, or both as the destination.
• After launch, track traffic, latency (response delay) and error rates (frequency of failures), and retain brand controls including verified logos and authenticated accounts (accounts that preserve a developer's verified identity).

Google says the setup lets a developer “connect their app to Google’s services” so it can be used inside the company's AI features. The choice of one surface or both gives developers control over distribution, although Google says nothing about how apps will be ranked.

Takeaway: Google is opening a submission and monitoring pipeline, but it has not set a general release date.

What changes for people using Gemini or Search?

For users, Google presents connected apps as a way to reach familiar services through an AI assistant. Instead of leaving the assistant to open a separate app, the user requests a task inside the linked service. In Google's design, the model supplies the interface and the connected service performs the action.

Google says a user could connect Canva to design a flyer or YouTube Music to build a playlist. The announcement provides no completion rates, usage volumes, or figures showing how often either connected app completes those tasks.

Google Search currently has fewer integrations than Gemini, including Instacart, Canva, YouTube Music and Yahoo Fantasy Football. Google named meditation app Calm as another partner, with its app due to launch on both Gemini and Search, but supplied no launch date.

The company says breadth is a condition for consumer adoption. That makes the number of submitted and live apps more relevant than the novelty of any single integration.

Takeaway: Connected apps make Gemini and Search action-oriented, but the announced integrations do not yet demonstrate broad adoption.

How large is the initial catalog?

Google says Gemini already has over 50 Connected Apps. Named third-party partners include Adobe, Experian, Localiza and Peloton, while Google's own catalog includes Workplace, Photos, YouTube Music and Google Home.

Google says those apps will come to Google Search over the next few weeks. The announcement does not provide a separate total for Search, so the size of its catalog before the broader Gemini rollout remains unspecified.

The gap gives Google a practical test: a large Gemini lineup must transfer into a Search surface that currently starts with fewer integrations. It also can leave availability uneven, because developers can support Search, Gemini or both rather than being required to support every surface.

Google disclosed no user-adoption figures, no split between its own apps and partner apps, and no named developer or customer discussing the earlier release. The numbers therefore describe catalog size, not demonstrated engagement.

Takeaway: Gemini has the larger catalog, while Search begins with fewer integrations and a gradual rollout.

What can developers measure and control?

After an app goes live, Google says developers can track traffic, latency and error rates. They can also maintain brand elements such as verified logos and authenticated accounts, preserving their identity inside Google's experience.

Google compares the opportunity to search engine optimization (SEO), the system companies use to rank websites in search results. The company says this could be an early step for businesses seeking visibility as users move to the “agentic web,” meaning AI interfaces that complete tasks across connected services.

Control over traffic and branding makes the offering more than a directory in Google's description. Yet the announcement does not explain how Google ranks competing apps, whether developers can influence results, or whether placement is paid.

Google also gives no pricing, revenue-share terms, review standards, technical detail behind its claim that connections are secure, or date for broad availability. A U.S. waitlist and limited preview therefore signal controlled testing rather than unrestricted platform access.

Takeaway: Google gives developers visibility and brand controls, but leaves key commercial and ranking terms undisclosed.

How strong is the agentic-web case?

Google's strongest evidence for the agentic-web thesis is practical rather than measured: Gemini already lists more than 50 connected apps, and Search already supports several familiar services. The announcement offers no data showing that people complete tasks more often through an AI interface than through the underlying app.

Google says ChatGPT and Meta’s Muse have taken varying approaches to consumer-facing integrations. It also states that an AI provider needs a wide range of connected apps to capture and retain attention, which makes developer recruitment as important as consumer demand in Google's own premise.

Because developers can support Search, Gemini or both, Google has created a path toward a broad catalog without guaranteeing identical availability on each surface. Google does not explain how it will handle duplicate services or whether consumers will see the same recommendations across the two products.

The announcement does not quote an executive, customer or independent developer, and it provides no case study, benchmark or revenue projection. That leaves the central commercial claim—businesses will value distribution through Gemini and Search enough to build for it—supported mainly by Google's product roadmap.

Takeaway: The platform has breadth, but the announcement does not yet demonstrate demand, engagement or a durable business model.

What should developers and users watch next?

Google's announcement sets three observable milestones. Each one tests a different part of the platform's rollout.

• The more than 50 connected apps currently in Gemini are scheduled to arrive in Search over the next few weeks.
• Calm is expected on both Gemini and Search, though Google gave no launch date.
• U.S. developers on the platform waitlist can register, test and submit apps, but Google has not said when the public preview will widen.

The first test is distribution, not novelty: developers need an audience, and users need enough integrations to try more than one service. Search's smaller starting catalog means Google must add depth while porting the existing Gemini lineup.

For businesses, the platform becomes consequential if visibility through AI starts to resemble search placement rather than a niche feature. For users, the next evidence will be whether the connections stay available, reliable and under each provider's brand control.

Takeaway: Google's next proof point is a fuller Search catalog, reliable connections and evidence that developers and users will adopt them.