Muse Code enters a market that Anthropic’s Claude Code and OpenAI’s Codex already dominate. Unlike Claude Code or Codex, which offer desktop apps or IDE integrations, Muse Code runs entirely through the command line. Artificial Analysis scores Muse Spark 1.2 at 54 on its Intelligence Index, up from 51 for Muse Spark 1.1 and 43 for the original Muse Spark in April. Unlike Claude Code (which requires a $20 Pro or $100-$200 Max plan) or Codex (bundled with ChatGPT plans), Muse Code is purely pay-as-you-go. Meta built Muse Spark 1.2 and Muse Code together, using rejection-sampled trajectories from the agent itself to improve the model.