New AI Assistants Again? Deciding Whether to Spend Time on Muse Code and Muse Spark

Meta has entered the AI coding race with two new products: Muse Spark 1.2, a next-generation reasoning model specialized for programming, and Muse Code, a coding agent that works on top of Muse Spark 1.2. The benchmarks for Muse Spark 1.2 place it second on Terminal Bench, behind Claude Opus 5, but the article claims that Meta omitted Chinese competitors like Alibaba's Qwen3.8-Max, because including them would drop Muse Spark to third place. The model is designed for autonomous agents, featuring multi-agent orchestration, parallel execution, computer control, GitHub agent creation, and search integration. Meta introduced a unique monetization strategy: the standard muse-spark-1.2 version costs 20 times more than the Contributor version, which offers a discount in exchange for allowing Meta to use the user's data for training. Muse Code is an agent that can plan and execute tasks, work with repositories, run terminal commands, write tests, and manage large codebases with a 1-million-token context window. The combination of Muse Code and Muse Spark aims to fully automate the development cycle, from task creation to testing and debugging.