Meta plans to launch Hatch AI agent platform in coming weeks

Meta ↗ 1.07% is preparing to release a consumer AI agent platform codenamed "Hatch" within the next several weeks, as the company moves to turn its massive artificial intelligence investments into direct revenue, The Information reported Monday, citing internal documents.investing+1

From Research to Revenue

Hatch is the consumer-facing version of Meta's internal OpenClaw AI agent and represents a core piece of CEO Mark Zuckerberg's strategy to diversify beyond advertising income. The platform is trained to navigate and interact with external web services including DoorDash ↗ 1.47%, Etsy ↘ 3.19%, Reddit ↘ 4.36%, Yelp ↘ 4.51%, and Microsoft ↗ 0.95% Outlook, according to the internal documents. Prototype designs feature a customizable dashboard with skills such as automated fitness tracking and travel itinerary planning.stocktwits+2

To monetize the technology, Meta has considered a tiered pricing structure with a premium subscription costing up to $199.99 per month for expanded usage capacity. That price point would place it alongside high-end AI subscriptions from competitors like OpenAI and Anthropic.investing+2

WhatsApp Integration and Watermelon Model

Alongside Hatch, Meta is preparing a WhatsApp integration that would allow users to connect with third-party AI agents directly within the messaging app, with a limited user trial slated to begin as early as this week. The company is also targeting an October release for its latest AI model, codenamed "Watermelon," though internal documents did not specify whether it will join an existing model line or launch as a standalone system.stocktwits+1

A Meta spokesperson declined to comment on the internal documents.stocktwits

Competitive Landscape

Meta enters an increasingly crowded market for agentic AI, where OpenAI and Anthropic have been building similar capabilities. What Meta brings is distribution — potential integration across Instagram, Facebook, WhatsApp, and Messenger could accelerate adoption far beyond what a standalone product might achieve. The company has also continued its open-source efforts, recently releasing Muse Glimmer and committing to an open-weight version of Muse Spark 1.2, maintaining a parallel strategy of commercial agents and open models to preserve developer loyalty.cryptobriefing+1