AI agents are starting to move out of the demo stage and into ordinary errands. That is the real story behind the new StepFun and Alipay partnership.
At the 2026 World Artificial Intelligence Conference in Shanghai, foundation model company StepFun and Alipay announced a system-level partnership designed to connect AI agents with real-world daily services. Not just answering questions. Not just summarizing a page. The goal is to let users ask for something in plain language and have the agent coordinate the steps across services.
That changes the pitch around AI agents a bit.
For years, people heard that assistants would “book things,” “find things,” and “handle tasks.” Most of the time, they still needed the user to jump between apps, approve every step, or manually complete the boring part. StepFun and Alipay are now trying to make that process feel more connected.
StepFun and Alipay Announce System-Level AI Agent Partnership
The partnership connects StepFun’s AI-native terminal and large model capabilities with Alipay’s AI-powered daily life services. Users will be able to access those services through Alipay’s Ah Bao by speaking to StepFun’s AI agent.
Ah Bao is the AI-native version of the Alipay app. It was launched in June 2026 and allows users to access more than 10,000 everyday services through natural conversations. These services include utility payments, pet service bookings, EV charging, and other daily-life tasks.
That number matters because AI agents need more than clever language skills. They need access to services that people actually use.
A chatbot can tell someone where an EV charger might be. A connected agent can help find one, compare options, and prepare the next step. That is where Alipay becomes important in this partnership.
Amoo Shows How the Partnership Works
During a live demonstration at WAIC 2026, StepFun showed Amoo, an AI agent running on its agentic smartphone StepX Neo.
The user gave a simple request: find the nearest EV charger and order coffee to their home.
Amoo then interpreted the request, considered route and price, and activated the relevant charging and coffee agents on Alipay’s Ah Bao platform. The system recommended options and generated an order that was ready for the user to confirm.
Small example. Big implication.
The point was not coffee. It was coordination. One request moved across different services without the user opening several apps and repeating the same details again and again.
This is exactly the kind of friction AI agent companies keep promising to remove.
Alipay’s Agent Hub-Access Powers the Connection
The collaboration is powered by Alipay’s Agent Hub-Access, also known as AHA. The system allows AI agents to connect, collaborate, and complete tasks across platforms with user authorization.
That last part is important.
When AI agents start touching payments, bookings, location, utilities, transport, and personal services, authorization cannot be treated like a small detail. Alipay has said that payments and other financial actions still require user confirmation, which makes sense. Nobody wants an AI assistant accidentally buying things in the background.
So the partnership is not about removing the user completely. It is about removing unnecessary app-hopping while keeping the final decision with the user.
That balance will probably decide how fast people trust agentic services.
Why This Matters for the AI Agent Race
A lot of AI agent announcements still feel vague. They talk about future productivity, future workflows, future automation. StepFun and Alipay are aiming at something more concrete: daily consumer services.
That makes the partnership interesting.
China’s AI market is moving quickly toward agents that can act across platforms. Alipay already has a massive service ecosystem. StepFun brings the foundation model and AI-native terminal side. Put those together, and the agent is no longer just a voice interface. It becomes a front door to services.
This is also where AI agents may become normal for non-technical users.
Most people do not care about benchmarks. They care if the tool can help them pay a bill, book a service, find a charger, order something, or complete a task faster than tapping through five screens.
That is the boring version of AI. It may also be the version that gets used the most.
Ah Bao Is Becoming an AI Agent Hub
Alipay’s Ah Bao is not only working with StepFun. The platform is already being integrated by other AI terminals and partners, including Qwen and OPPO.
On July 15, 2026, OPPO partnered with Alipay to let its Breeno AI agent access Ah Bao’s daily-life services through conversation. That gives Alipay a broader role in the emerging agent ecosystem.
In other words, Alipay does not appear to be building Ah Bao as a single app feature. It is shaping it into a service layer for different AI agents and devices.
That could become a strong position if agentic smartphones, AI assistants, and model-powered devices continue to grow. Whoever controls access to daily services may control a major part of the user experience.
StepFun Gets a Real-World Use Case for Its AI Terminal
For StepFun, the partnership gives its StepX Neo device and Amoo agent something practical to do.
That matters because AI-native devices have struggled with a simple question: why should people use them instead of a normal smartphone app?
By connecting Amoo to Alipay’s service ecosystem, StepFun can show a clearer use case. The phone is not just running an AI model. It is trying to become a task-completion device.
The challenge, of course, is execution. AI agents need to understand messy human requests, call the right services, avoid mistakes, protect user data, and still feel faster than doing it manually.
That is a lot to get right.
AI Agents Are Moving From Conversation to Action
The StepFun and Alipay partnership shows where consumer AI may be heading next. The industry is moving from chat-based answers toward agents that can carry out connected actions.
Not fully autonomous. Not magic. But more useful than a chatbot sitting in a separate window.
The real test will be whether users actually adopt this behavior in daily life. Asking an AI agent to find a charger and order coffee sounds convenient. Trusting it with payments, bookings, routes, preferences, and repeated errands is another step.
Still, the direction is clear.
AI agents are no longer being judged only by how well they talk. They are being judged by what they can actually get done.
