Meta buys Conductor, the Mac app that runs Claude Code and Codex agents in parallel
Terms were not disclosed. Conductor Cloud goes read-only on 13 November and shuts on 20 November, and the local app stops receiving updates in early 2027 while the team joins Meta.
Meta has acquired Conductor, a developer-tools start-up whose Mac application lets one engineer run several coding agents at the same time, each in its own copy of the repository, then review and merge the results. Axios first reported the deal on 4 November 2026 and Conductor confirmed it on its website. Terms were not disclosed. SiliconANGLE reports that the company, legally Melty Inc., was founded in 2024 by Charlie Holtz and Jackson de Campos, raised a $22 million Series A in March from Spark Capital and Matrix, and says more than 100,000 people use the product. Supported agents are Anthropic’s Claude Code, OpenAI’s Codex and Cursor, paid for through users’ own subscriptions. Engineers at Meta were already using it.
The product is being wound down rather than absorbed. New sign-ups and downloads have stopped, Conductor Cloud becomes read-only on 13 November and shuts on 20 November, after which stored cloud data is deleted and active subscribers are refunded. The local app keeps receiving new models and critical fixes until early 2027. Conductor says a new interface is planned for early 2027 but has not described it. InfoWorld, writing on 5 November, quotes Omdia analyst Lian Jye Su saying Conductor’s advantage is that it “sits above individual coding agents”. Neither company has said whether the Claude Code and Codex integrations will survive, and Meta has made no statement of its own. SiliconANGLE places the deal alongside Meta’s October releases of Muse Spark Open and the closed Muse Wonder models.
Why it matters
If you rely on Conductor Cloud, export or migrate workspaces before 13 November, when it goes read-only. The wider signal is that the orchestration layer above coding agents is being bought rather than built, and that a model vendor now owns a tool whose appeal was being vendor-neutral. Treat any multi-agent tool that depends on one lab’s goodwill as a dependency with a shelf life. The InfoWorld analysts’ point stands regardless of owner: human review before merge is the control that matters when agents work in parallel, and agent-to-agent communication widens the attack surface.