Meta acquires AI coding startup Conductor to gain multi-agent orchestration

Meta has acquired Conductor, an AI coding startup that lets users spin up software projects by deploying multiple agents into parallel workspaces. Meta confirmed the acquisition to TechCrunch but declined to comment further or disclose the terms. The news was first reported by Axios, and Conductor announced the move on its own blog the same day.

Conductor founder and CEO Charlie Holtz framed the deal as a platform shift rather than an exit. "I have big news: Conductor is joining Meta," he wrote on Wednesday, adding that the company has "big, crazy plans for the next interface" and would have "something new" for customers to use by early next year.

Holtz also said he was looking forward to working with Alexandr Wang, Meta's chief AI officer, and Nat Friedman, who leads product at Meta's Superintelligence Labs. In a post on X, he said the two have been "on an absolute tear" and argued that MSL has "the highest talent density of all the labs." That kind of language usually signals the team is being absorbed into the parent organization rather than kept as a standalone unit.

What Conductor actually sells is a way of working, not a model. Founded in 2024, it only launched its core product last summer. Its agentic teams can be deployed in the cloud, locally, or through a user's phone, and are designed to split a project into separate, compartmentalized parts that different agents handle in concert. The agents themselves are not built in-house; they come from major labs including Anthropic and OpenAI, which leaves Conductor responsible for orchestration, context allocation and stitching results together.

The transition plan reinforces that reading. Holtz said Conductor's core features will keep working as they do now until early next year, with one exception: the company's cloud service shuts down on November 20, after which it will begin deleting user cloud data. For existing customers that is both a grace period and a hard deadline to move workflows back to local machines or to another provider.

For Meta, the acquisition looks less like a coding-tool purchase than a bet on orchestration: the layer that decides how several agents split a job, not the models that actually do the work. Meta has been moving quickly on coding agents, launching its own Muse Code in August, reporting more than 10 million users across its Muse agents, and picking up properties such as the viral agent platform Moltbook. Folding Conductor's parallel-workspace approach into Superintelligence Labs buys engineering experience in multi-agent coordination that is hard to replicate from scratch.

Three things are worth watching. The deal terms remain undisclosed, and it is still unclear in what form the team joins MSL. Holtz's promised "next interface" probably will not surface until early next year. And Conductor's reliance on Anthropic and OpenAI models raises an obvious question: once inside Meta, does that stack get swapped for in-house models? The answers will determine whether this was a product fill-in or a change of direction for Meta's coding agents.