Meta Launches Muse Code to Challenge Claude Code and OpenAI Codex
Meta launched Muse Code, a beta AI coding agent powered by Muse Spark 1.2 that can write, debug and verify code while managing longer software-development assignments.
Why it matters: Meta is moving directly into the AI coding-agent market, where developer adoption, pricing and workflow reliability could shape enterprise AI spending.
Meta has launched Muse Code, a new artificial-intelligence agent designed to write, test and debug software.
The beta product is powered by Muse Spark 1.2, Meta’s latest model for coding and agent-based work. Meta said it trained the model and coding agent together so they could operate as an integrated system.
Muse Code is designed to complete extended software-engineering assignments rather than only suggest individual lines of code. It can divide difficult work among several sub-agents, verify results and maintain a persistent activity history.
That history allows the agent to resume work following a crash or interruption without repeating the entire assignment.
Meta is positioning the product against Anthropic’s Claude Code and OpenAI’s Codex, two of the leading tools in the rapidly expanding AI-coding market.
The standard pay-as-you-go plan costs $1.25 per million input tokens and $4.25 per million output tokens. Meta is using aggressive pricing to attract developers and increase usage of its wider model platform.
Coding has become one of the clearest commercial markets for AI agents because performance can be measured through completed features, resolved defects and functioning software.
Muse Code gives Meta a potential route into enterprise-development workflows, but its competitive position remains unproven. Independent evaluations will be needed to determine whether its low price is accompanied by the reliability required for production codebases.