Meta Looks to Charge Up to $200 a Month for Planned ‘Hatch’ AI Agent
The Takeaway
- Meta’s planned Hatch AI agent could cost up to $200 monthly, rivaling top AI subscriptions.
- Hatch will create software tools and handle tasks such as managing schedules and sending emails for users.
- A premium version of Hatch would be part of Meta’s efforts to diversify revenue beyond advertising.
Meta Platforms is considering charging up to $200 a month for its planned consumer version of the OpenClaw AI agent tool, according to internal documents, an ambitious target that would rival top tier offerings from established AI giants.
The product, currently called Hatch, could launch with tiered pricing including the $199.99 premium subscription that would include higher usage limits, according to the documents reviewed by The Information and a person familiar with the matter, who noted that final pricing decisions haven’t been made.
Hatch would handle a range of tasks from vibe coding new software tools to scheduling events on users’ calendars or sending emails on their behalf, according to images of test versions also viewed by The Information. Users can describe what they want in plain language, similar to the way vibe coding apps work, such as “build me a fitness tracker,” and Hatch generates a working tool.
Charging for a premium version of Hatch would place Meta in more direct competition with OpenAI, Anthropic and other AI firms that already have built large customer bases for their AI coding and agent products. Both OpenAI’s ChatGPT Pro subscription and the highest-end version of Anthropic’s Claude Max plan cost $200 per month.
As previously reported, Meta is conceiving of Hatch as essentially a consumer version of OpenClaw, the open-source agent tool that has become popular in tech circles but complicated to use. The planned product is part of Meta Chief Executive Mark Zuckerberg’s broader effort to develop new revenue streams to help pay for its enormous investments in AI infrastructure and model development.
The documents, which date from April and May, show that Hatch also includes a customizable feed or dashboard to display the tools created by the agent, similar to OpenClaw, which lets users pick from capabilities called “skills” where modular add-ons let AI agents access tools, automate workflows, and interact with third-party services. These skills will allow Hatch to carry out specific tasks, such as creating a travel itinerary, according to the person familiar with the matter.
Meta-Powered AI
In its development so far, Hatch has been powered by Anthropic’s Claude Opus 4.6 and Claude Sonnet 4.6 models. When launched, Hatch will be powered by Meta’s latest AI model, Muse Spark, the person said.
Upgrade to ask Deep Research to…
What risks surround SpaceX’s massive Anthropic deal?
What threatens startups using Anthropic and OpenAI?
What could derail the 2026 AI IPOs?
How are AI costs hurting SaaS vendors?
In an April internal memo, product marketing manager Jennifer Lin said Meta was working to expand the agentic capabilities of skills, but that no final decisions had been made on branding, and that the name “Meta AI Skills” had not been confirmed.
The internal documents indicate Meta was initially targeting a U.S. launch in April. A separate internal memo from early May said Meta was aiming for a controlled rollout to about 10 businesses to pilot the product, followed by a broader launch with finalized branding and product naming in July, although the launch date could change.
In one of the internal documents, Meta describes plans to launch a paid premium subscription tier, “Hatch Plus,” which would offer significantly higher usage limits than the free version—between five and ten times more daily capacity. The document also states that token allowances would reset each billing cycle and would not roll over, “similar to Claude, OpenAI, etc.”
Zuckerberg has said that Meta intends to build AI agents for business as well as individual users, whom he sees as central to his vision for what he calls personal superintelligence. “Our goal is not just to deliver Meta AI as an assistant, but to deliver agents that can understand your goals and then work day and night to help you achieve them,” he said in Meta’s second-quarter earnings call in April.
The Meta chief also dropped hints about users being able to vibe-code their own tools, saying that whether people use its personal or business agents to achieve their goals, he believes there will be a “massive increase in entrepreneurship from people creating new things that they’ve always wanted to exist but previously didn’t have the tools to bring into the world.”
Meta has increasingly focused on turning its AI investments into revenue-generating products as the company continues to invest heavily in infrastructure, chips, and model development. Its AI capabilities helped drive strong growth in 2025, when revenue topped $200 billion. But almost all of that still comes from advertising, and Meta is under investor pressure to add new revenue streams to justify its capital expenditures, which it has said could reach $145 billion this year, about double the level in 2025.
The company has already begun rolling out paid features for its AI offerings, including subscription plans tied to its Meta AI chatbot, which is accessible through Facebook, Instagram and WhatsApp. This week, Meta also said it is expanding existing AI agent tools that businesses can use within its platforms, extending its Meta Business Agent product to Instagram and launching Meta Business Agent Platform.
The agents in those tools can recommend products, close sales, book appointments, answer customer questions, and send personalized marketing messages, as well as provide business owners with overnight chat summaries and performance insights. The Meta Business Agent tools are currently free, and Meta said it plans to introduce paid subscriptions in the coming months. Those tools are intended for use on Meta’s own platforms, whereas Hatch will also be able to function in the wider digital world.