Meta Platforms is preparing to launch Hatch, a consumer-focused AI agent platform, in the coming weeks — and plans to follow it with a new frontier model codenamed Watermelon in October, according to internal documents seen by The Information. The launch would be Meta's most aggressive attempt yet to turn its enormous AI investment into subscription revenue, with a premium tier reported to run as much as $199.99 per month. For readers tracking the consumer AI tools race, Hatch is shaping up to be one of the biggest launches of the fall.
The plans, reported by The Information and confirmed in detail by The Decoder, describe Hatch as a consumer version of Meta's OpenClaw agent — the company's internal agentic AI system — repackaged for everyday users rather than developers.
What Hatch Will Actually Do
According to the internal documents, Hatch is designed to connect directly to a range of popular services and platforms, including DoorDash, Etsy, Reddit, Yelp, and Microsoft Outlook. Rather than simply answering questions in a chat window, the agent is built to take actions across those services on the user's behalf — ordering food, shopping, researching recommendations, and managing communications.
Hatch will also offer a customizable dashboard, the documents indicate, with modular features such as a fitness tracker and a trip planner. The vision described is closer to a personal operations hub than a chatbot: a persistent interface where users can see and manage the tasks their AI agent is running across the apps they already use.
On WhatsApp, Meta is separately building a platform that will let users plug in additional AI agents beyond its own — a move that would open the company's massive messaging base to third-party agentic services.
Premium Pricing, Big Ambitions
The most attention-grabbing detail is the price. Meta is reportedly weighing a tiered pricing model for Hatch, with a premium subscription running up to $199.99 per month. That would put Meta's top tier at the very top of the consumer AI market — well above typical premium AI subscriptions, and squarely in the territory usually reserved for professional-grade software.
The pricing logic reflects Meta's broader strategic problem. The company has spent tens of billions of dollars building out AI infrastructure and its Llama-family models, but nearly all of its revenue still comes from advertising. CEO Mark Zuckerberg has increasingly framed consumer AI products as the path to diversifying that revenue — and justifying the capital expenditure to investors.
Hatch, in other words, is not just a product launch. It is the monetization strategy. An agent that books, buys, plans, and coordinates is a product people might plausibly pay subscription fees for — and one that generates high-value engagement data that could also feed Meta's advertising machine.
Additional reporting on the launch timeline suggests Meta is targeting early September for the initial Hatch rollout, which would put the platform in front of users well before the holiday shopping season — a window in which a shopping-and-planning agent could prove its worth quickly. A staggered rollout, starting with a subset of users in the United States before a broader international push, remains likely given the number of third-party integrations involved, though Meta has not confirmed launch details publicly.
Watermelon: The Model Beneath the Agent
The October-bound model, codenamed Watermelon, is the other half of the equation. According to The Information's reporting, it remains unclear whether Watermelon will become part of Meta's Muse model family or ship as a standalone model.
What is clear is that consumer agents are only as good as the models powering them. Agentic workloads — multi-step reasoning, tool use, reliable instruction-following over long sessions — demand capabilities that current models still struggle with. A purpose-built fall model release timed alongside Hatch's rollout suggests Meta intends to iterate the model and the product together, in the same full-stack manner that rivals like OpenAI and Google have adopted.
A Crowded Field, A Familiar Playbook
Meta enters a consumer agent market that is heating up quickly. OpenAI, Google, and a wave of startups are all racing to ship agents that can complete real tasks rather than just converse. What Meta brings to the table is distribution few rivals can match: billions of users across Facebook, Instagram, and WhatsApp, and deep existing integration with the services people use daily.
The risk is equally familiar: consumers have historically shown limited willingness to pay for AI features, and the industry is littered with agent launches that overpromised. A $199.99 price point will invite intense scrutiny of what Hatch can actually deliver at launch. Privacy questions will follow too — an agent that reads your Outlook, orders your DoorDash, and browses your Reddit will know a great deal about you.
Neither Meta nor The Information has publicly commented beyond the reporting, and plans for launch timing and pricing could still change before the public debut. But with Hatch reportedly weeks away and Watermelon following in October, Meta's answer to the question of how it makes money from AI is about to be put to a very public test.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
