Meta Launches Muse Code AI Agent
Meta has introduced Muse Code, its first AI coding agent, entering a market led by Anthropic’s Claude Code and OpenAI’s Codex. Currently available in beta, the terminal-based tool is powered by Meta’s latest Muse Spark 1.2 model and targets professional software developers working on large codebases.
Unlike conventional AI coding assistants, Muse Code manages complex software engineering tasks across entire repositories. It can plan code changes, generate new code, execute tests, validate results, and coordinate multiple AI sub-agents simultaneously. As a result, developers can automate larger portions of the software development lifecycle from a single interface.
Built for large-scale software engineering
Muse Code operates directly from the command line and supports long-running development workflows. Moreover, the agent creates isolated workspaces for individual tasks, allowing multiple sub-agents to work in parallel without affecting the original codebase. It also maintains a persistent activity log, enabling projects to resume after interruptions without restarting completed work.
Mark Zuckerberg described the tool’s capabilities in a public post. “Muse Code, which is currently available in beta, can handle full software engineering tasks across large repositories.”
He also noted that larger projects automatically split into parallel sub-agents working independently before merging their results.
Competing on price and developer adoption
Meta has adopted a pay-as-you-go pricing model for Muse Code instead of a monthly subscription. Standard pricing starts at $1.25 per million input tokens and $4.25 per million output tokens. Additionally, the company offers a lower-cost contributor tier for developers who allow Meta to use their prompts and code completions to improve future AI models.
The launch marks Meta’s latest effort to expand beyond foundation models into AI-powered developer tools. Consequently, Muse Code now competes directly with Anthropic’s Claude Code and OpenAI’s Codex in one of the fastest-growing segments of enterprise AI. While affordability may attract startups and independent developers, Meta will also need to demonstrate strong performance and reliability to gain wider enterprise adoption.
