Meta Platforms is preparing a paid AI agent called Hatch for launch in the coming weeks, while a new AI model named Watermelon is expected in October. The plans point to a broader effort inside Meta to package AI into consumer-facing products that can generate revenue beyond advertising.
What Hatch Is Expected To Do
Hatch is set to become a consumer version of the OpenClaw agent. According to internal documents seen by The Information, the product is part of CEO Mark Zuckerberg's plan to make money from Meta's heavy AI spending and expand the company's revenue mix.
The agent is designed to connect with outside sites and services, including DoorDash, Etsy, Reddit, Yelp, and Outlook. That list suggests Hatch is meant to do more than answer questions. It is being positioned as a tool that can sit across everyday online tasks, from planning to shopping to organizing information.
Meta is also preparing a customizable dashboard for Hatch. The source describes possible features such as a fitness tracker or trip planner, which would make the agent feel less like a single chatbot window and more like a personal workspace shaped around a user's needs.
Pricing Could Reach A Premium Tier
Meta is weighing a tiered pricing model for Hatch. The source says one premium subscription could run up to $199.99 a month.
That figure matters because it places Hatch in a different category from casual AI tools. A high-end tier would imply that Meta sees at least some users paying for broader agent capabilities, deeper integrations, or more advanced features. The source does not specify what each pricing level would include, so the exact value proposition remains unclear.
Still, the basic direction is visible. Meta is exploring paid AI services at a time when it wants returns from major AI investment. Hatch appears to be one of the clearest consumer-facing pieces of that strategy.
Watermelon Adds Another Piece To Meta's AI Roadmap
Alongside Hatch, Meta plans to release a new AI model called Watermelon in October. The source does not say whether Watermelon will become part of the Muse model family or ship as a standalone model.
That uncertainty leaves an important product question open. If Watermelon is folded into Muse, it may serve as an extension of an existing model line. If it ships separately, Meta may be preparing a distinct model identity for different capabilities or use cases.
The source also does not say whether Watermelon will directly power Hatch. For now, the two items sit side by side in Meta's reported AI plans: a paid agent arriving soon, and a new model coming in October.
Why The WhatsApp Platform Matters
Meta is also building a platform on WhatsApp that will let users plug in more AI agents. That effort suggests Meta is not only thinking about one assistant, but about an environment where multiple agents can be added and used inside an existing messaging product.
This matters because agents become more useful when they can meet users where they already spend time. WhatsApp could give Meta a familiar surface for AI agent access, while Hatch could provide a more customized consumer experience through its dashboard and service connections.
The source does not describe how the WhatsApp platform will work, what agents will be available, or whether it will connect to Hatch. But its presence in the same set of plans shows that Meta is building toward a broader agent ecosystem rather than a single isolated launch.
The Bigger Direction
The reported Hatch launch, the Watermelon model release, and the WhatsApp agent platform all point in the same direction: Meta is trying to turn AI development into products users may pay for or use regularly.
Several facts are still unresolved. Hatch's final pricing structure has not been detailed. Watermelon's place in Meta's model lineup is unclear. The WhatsApp platform has been described only in broad terms. But the overall shape is becoming easier to read.
Meta is preparing AI agents that connect to popular services, live in customizable interfaces, and potentially expand through messaging. If Hatch launches in the coming weeks as described, it will become a visible test of whether Meta can move its AI work from infrastructure and models into paid consumer products.