Meta Prepares Hatch AI Agent and Watermelon
Meta Platforms is reportedly preparing two major AI launches as it seeks to turn its growing investment in artificial intelligence into new consumer revenue. The company plans to release a consumer AI agent platform called Hatch within the coming weeks, according to The Information.
Meanwhile, Meta is reportedly targeting October for a new AI model internally known as Watermelon. The plans highlight CEO Mark Zuckerberg’s push to expand Meta’s AI business beyond advertising.
Hatch Targets Consumer AI, Agents
Hatch is reportedly based on Meta’s internal OpenClaw technology. The platform could allow an AI agent to interact with external websites and services while completing tasks for users.
Initially, the platform may focus on consumer use cases such as shopping. However, its broader purpose appears to involve autonomous task execution across online services. That would move Meta AI further toward agentic computing.
Meta has already expanded its AI assistant beyond conventional question-and-answer interactions. In July, the company said Meta AI could create plans, connect with email and calendar applications, conduct research and complete tasks for users.
Furthermore, Meta says it wants every person to have a personal AI agent capable of working on their behalf. The company envisions agents handling areas such as finances, home management, careers and other personal tasks.
The reported Hatch launch therefore fits into a wider product strategy. Instead of limiting AI to conversations, Meta is moving toward software that can take actions across connected services.
Meta is also reportedly considering a tiered subscription model for Hatch. The highest-priced plan could cost as much as $199.99 per month, although the pricing remains under consideration.
Watermelon Model Targets Frontier AI
Alongside Hatch, Meta is reportedly preparing Watermelon, a new AI model expected in October. The Information reported the planned timing based on internal documents.
However, public information about Watermelon remains limited. Meta has not formally announced the model or published independent benchmark results. Therefore, claims about its performance should be treated as reported rather than established results.
The model also arrives as Meta rebuilds its AI strategy around personal superintelligence. The company says it plans to develop leading models while making AI widely available through its products and services.
Meanwhile, Meta continues to develop models and agentic capabilities across its consumer ecosystem. Its current Meta AI experience can already plan tasks, connect to applications, and conduct research.
Watermelon could consequently become the model layer supporting a more capable generation of Meta AI products. Yet, its actual capabilities will only become clear after Meta releases technical information and independent evaluations.
The timing is also significant because Meta faces intense competition from OpenAI, Google and Anthropic. Therefore, a stronger model could support the company’s broader effort to build a competitive consumer AI platform.
Meta Seeks New AI Revenue Streams
The reported Hatch strategy comes as Meta looks for ways to monetise its substantial AI investment. Advertising remains the company’s core revenue engine, but AI could create additional subscription and commerce opportunities.
Meta has already begun building commercial AI products. Its Meta Business Agent allows companies to answer customer questions, recommend products, book appointments and close sales. The company is also offering a platform for businesses to build and deploy agents at scale.
Moreover, Meta says paid subscription offerings for Business Agent will expand in the coming months. That provides an existing example of how the company intends to attach revenue models to agentic AI.
Hatch could extend that approach to consumers. A subscription model would give Meta a direct revenue stream from advanced AI capabilities, while agent-driven shopping could create additional commercial opportunities.
At the same time, Meta is positioning personal AI agents as a major part of its long-term strategy. Zuckerberg has said the company aims to deliver personal superintelligence to billions of people and small businesses.
The reported Hatch and Watermelon launches therefore represent more than two product developments. Together, they point toward Meta’s attempt to combine stronger AI models with agents that can act across digital services.
However, execution will determine whether that strategy becomes commercially significant. Hatch must prove that consumers will trust agents with meaningful tasks, while Watermelon must demonstrate competitive performance outside internal testing.
For now, both remain reported developments rather than fully launched products. Consequently, Meta’s next releases will provide the clearest evidence of whether its AI investments can translate into a scalable consumer business.
