Meta Launches Enterprise Platform to Sell Muse AI Agent and Coding Tools to Businesses

Meta is creating a new business unit to sell AI services directly to companies, marking its first serious push into enterprise AI monetization. The unit, called the Meta Enterprise Platform, will offer the Muse agent, the Meta Business Agent, the Muse API, and Muse Code as its initial product lineup.

Chirantan Desai, former CEO of database company MongoDB, will lead the new business and report directly to Mark Zuckerberg. The appointment signals that Meta is treating enterprise AI as a priority separate from its consumer-facing products like the Meta AI app and Ray-Ban smart glasses.

Zuckerberg said Meta possesses advantages that few competitors can match, citing advanced models, leading agents, and massive infrastructure, according to Meta. The company has not disclosed pricing, API rate limits, or specific technical specifications for any of the four announced products. Meta also has not specified a launch date for general availability.

The Spending Pressure Behind the Move

The enterprise push comes as Meta faces mounting pressure to show returns on its AI infrastructure bet. The Wall Street Journal reports that Meta is spending more than $100 billion on AI infrastructure this year alone. That figure covers data centers, chips, and compute capacity built primarily to train and run Meta's own models, including its Llama family and newer Muse-branded products.

Until now, most of that spending has supported free consumer products and internal ad-targeting improvements rather than direct revenue streams. Selling Muse-based tools to businesses gives Meta a path to convert infrastructure costs into paying customer relationships, a model already proven by OpenAI, Anthropic, and Microsoft.

A Crowded Field

Meta enters an enterprise AI coding and agent market that is already dense with competition. Muse Code will compete directly with Anthropic's Claude Code, OpenAI's Codex, and Cursor, all of which have established developer followings and enterprise contracts. Meta also faces pressure from low-cost, open-weight models coming out of China, which have undercut pricing across the industry and made differentiation harder for any vendor charging premium rates.

Meta has not detailed how Muse Code or the Muse API differ technically from competing products, nor has it published benchmark comparisons. The company also hasn't said whether Muse products will run on Meta's own infrastructure exclusively or support deployment through third-party clouds, a factor that matters to enterprise buyers concerned about data residency and vendor lock-in.

What this means

Meta is late to enterprise AI monetization compared to OpenAI and Anthropic, both of which have run business-focused products for years. Bringing in a MongoDB veteran rather than an internal AI researcher suggests Meta sees this as fundamentally a sales and go-to-market problem, not a model-quality problem. That's a reasonable bet if Meta's underlying models are competitive, but the announcement contains no pricing, no benchmarks, and no technical differentiation versus Claude Code or Codex. Until Meta publishes those details, this reads as an organizational restructuring rather than a product with a clear reason for enterprises to switch.