Meta names five enterprise AI products without pricing or rollout

Meta has created Meta Enterprise Platform, a business unit that will sell its AI stack to companies and developers rather than rely solely on consumer-facing AI products and advertising. The effort starts with five named components — the Muse agent, Meta Business Agent, Muse API, Muse Code, and additional tools — but Meta has not disclosed pricing, availability dates, deployment models, service levels, or technical specifications for any of them.

The September 28, 2026 announcement formalizes Meta’s effort to turn AI infrastructure and agents into business-facing products. Mark Zuckerberg called it the company’s “next major pillar” and cited Meta’s advanced models, agents, infrastructure, and relationships with businesses as the offering’s foundation.

Meta has hired Chirantan “CJ” Desai as Chief Enterprise Platform Officer to run the unit. Desai joins from MongoDB, where he was CEO and President, and will report directly to Zuckerberg. Meta’s announcement of Meta Enterprise Platform says his prior roles included leading product and engineering at Cloudflare and nearly eight years at ServiceNow, including as president and COO.

“Meta Enterprise Platform will use our strengths that few other companies have: advanced models, leading agents, large-scale infrastructure, and years of working closely with many businesses.”

A product family, not yet a product catalog

Meta says its first focus is bringing its “full technology stack” to businesses and developers. It is not announcing one standalone enterprise assistant. The unit groups several parts of an emerging stack: an agent, a business-oriented agent, an API surface, and a code-oriented product.

Recommended reading

Microsoft’s Copilot redesign hides its real change: metered agents

Sergey Kuznetsov • Sep 25, 2026 • 13 min read

Muse is the clearest existing component. Meta introduced the personal assistant earlier in September 2026 with the ability to handle tasks such as sending email and booking travel. Meta Enterprise Platform makes Muse the lead product in a corporate package alongside the Meta Business Agent, Muse API, and Muse Code.

Meta has not said whether customers will consume Muse through a hosted API, deploy it in a private environment, connect it to their own business data, or buy it as a subscription. There are no published usage prices, context-window figures, model names, performance measurements, regional availability details, compliance certifications, or terms for handling customer data.

Meta says security and privacy are built into its enterprise products “from the outset.” The announcement does not specify identity controls, data retention, audit logging, encryption arrangements, tenant isolation, third-party integrations, or where customer prompts and outputs are processed. Those omissions leave the platform’s fit for regulated or security-sensitive deployments unresolved.

Desai’s move leaves MongoDB with an interim CEO

The leadership change is central to the launch. Meta is putting an executive with database, infrastructure, cloud, application, and security experience in charge of a platform it says will turn internal AI capabilities into business products.

MongoDB shares fell more than 17% after Desai’s departure was disclosed. The hire shows Meta views enterprise AI as an operating business requiring a dedicated executive rather than an extension of its research, advertising, or consumer-assistant groups.

“Meta Enterprise Platform will focus on turning its AI stack into products and services that companies can deploy for their own businesses.”

Desai’s phrasing is narrower than a promise of immediate general availability. “Products and services that companies can deploy” identifies the commercial goal, but Meta has not detailed which products are ready now, which customers will get them first, or what “deploy” means in practice.

The missing revenue model behind Meta’s AI spending

The new unit arrives as Meta tries to establish a business case for an AI buildout that Zuckerberg has said could reach $600 billion over the next two years, much of it directed at data centers. Enterprise software and API sales could give Meta a more direct way to monetize those investments than subscriptions aimed at Facebook and Instagram users, where broad willingness to pay has not yet been established.

The timing puts Meta into a more explicit contest for corporate AI budgets. Meta cites advanced models, agents, large-scale infrastructure, and its history of serving advertisers and businesses as advantages. The announcement does not include the procurement details enterprises use to compare platforms: predictable billing, contractual uptime, access controls, integration support, model choice, and data-governance terms.

Meta has reportedly considered leasing some of its data-center infrastructure to other companies, but it has not announced such a service. The September 28 announcement also does not say that Enterprise Platform will expose Meta infrastructure directly to customers. The confirmed scope is commercialization of Meta’s AI stack, not an infrastructure-leasing business.

Meta’s earlier enterprise foothold was adjacent to this effort

Meta had already been extending AI into business workflows before establishing this unit. In August 2026, its Mac application added screen sharing, system-wide dictation, Google Workspace access, business analytics, and recurring reports in Meta’s screen-aware Mac assistant. That direction suggested an interest in work-oriented use cases, but Meta Enterprise Platform creates a dedicated organization for selling technology to companies.

The shift comes as other AI companies seek revenue beyond consumer chat subscriptions. In August 2026, OpenAI said its enterprise revenue had overtaken consumer ChatGPT revenue and that its annualized run rate had reached $40 billion, ahead of its 2026 forecast. Meta has not provided a corresponding revenue target, enterprise customer count, or forecast for its new unit.

Meta is announcing an organizational and product strategy, not demonstrated enterprise demand or a booked-revenue business. It may be able to pair AI offerings with relationships built through advertising and business messaging. It must show that the named products can meet enterprise requirements and command prices that justify its infrastructure spending.

The announcement establishes intent, not commercial terms

The most concrete part of the launch is the leadership structure: Desai reports directly to Zuckerberg and owns a unit tasked with commercializing Meta’s AI stack. The product list establishes the intended scope across agents, APIs, code tooling, and business automation.

The buying details remain unpublished. There is no price sheet, customer list, release schedule, API documentation, or performance data. Meta has created the enterprise sales vehicle but has not shown the technical and commercial terms that will determine whether companies adopt it.