Muse Code is Meta's AI terminal coding agent, powered by Muse Spark 1.2. Parallel background agents, isolated git worktrees, a full event log, and a 1M-token context window.https://musecodes.io/enhttps://musecodes.io/blog/meta-enters-the-agent-race/https://musecodes.io/blog/meta-enters-the-agent-race/Strip away the benchmarks and Muse Code is a strategic statement: Meta’s first coding agent, priced to undercut the incumbents, shipped by the lab Alexandr Wang was hired to run.Fri, 07 Aug 2026 00:00:00 GMTmetastrategypricingagentshttps://musecodes.io/blog/muse-code-vs-the-field/https://musecodes.io/blog/muse-code-vs-the-field/Five serious rivals, two axes that matter. Where Meta’s agent actually sits against Claude, Codex, DeepSeek, GLM, and MiniMax — and a decision framework that survives the next price change.Fri, 07 Aug 2026 00:00:00 GMTbenchmarkspricingcomparisonagentshttps://musecodes.io/blog/agent-fanout-worktrees/https://musecodes.io/blog/agent-fanout-worktrees/Parallelism is easy to promise and hard to make safe. Muse Code’s answer is old, boring git technology used exactly right.Thu, 06 Aug 2026 00:00:00 GMTworktreesparallel-agentsgitmulti-agenthttps://musecodes.io/blog/async-background-agents/https://musecodes.io/blog/async-background-agents/The quiet architectural decision behind Muse Code’s speed: background agents that persist for the whole session instead of being spawned per task and thrown away.Thu, 06 Aug 2026 00:00:00 GMTarchitecturemulti-agentbackground-agentshttps://musecodes.io/blog/bundled-skills/https://musecodes.io/blog/bundled-skills/Meta’s one-line pitch for Muse Code’s bundled skills: put a rough idea in, get a grilled, taste-checked feature out. Here’s what each command actually does and when to reach for it.Thu, 06 Aug 2026 00:00:00 GMTskillsslash-commandsworkflowplanninghttps://musecodes.io/blog/contributor-vs-standard/https://musecodes.io/blog/contributor-vs-standard/The two ways to pay for Muse Spark 1.2 differ by more than an order of magnitude. The right choice isn’t one tier — it’s a per-project policy.Thu, 06 Aug 2026 00:00:00 GMTpricingprivacycost-optimizationhttps://musecodes.io/blog/cookbook-guide/https://musecodes.io/blog/cookbook-guide/Meta’s cookbook promises recipes that run the first time you copy them. The promise holds because of how it’s structured — each section assumes exactly what the previous one proved.Thu, 06 Aug 2026 00:00:00 GMTcookbookmodel-apirecipesagentshttps://musecodes.io/blog/event-log-replay/https://musecodes.io/blog/event-log-replay/Agents that work for hours need a memory that survives anything. Muse Code’s runtime is built around a single append-only log — and almost every trust property of the product falls out of it.Thu, 06 Aug 2026 00:00:00 GMTevent-logauditabilityreliabilityevent-sourcinghttps://musecodes.io/blog/getting-started-with-muse-code/https://musecodes.io/blog/getting-started-with-muse-code/Meta’s terminal coding agent is deliberately minimal to set up: one command, one browser sign-in, and you’re driving a multi-agent session inside your own repository. Here’s the whole path, including the parts worth slowing down for.Thu, 06 Aug 2026 00:00:00 GMTgetting-startedinstallclisetuphttps://musecodes.io/blog/github-agent/https://musecodes.io/blog/github-agent/The cookbook’s GitHub agent recipe wires triage, PR review, and bug-fix automation into GitHub Actions on OpenCode and Muse Spark. Its most transferable lessons are the guardrails.Thu, 06 Aug 2026 00:00:00 GMTgithub-actionsautomationagentscode-reviewhttps://musecodes.io/blog/model-api-ecosystem/https://musecodes.io/blog/model-api-ecosystem/Meta shipped its coding model everywhere developers already are — the OpenAI SDK, the Anthropic SDK, rival agent CLIs, even OpenRouter. Compatibility-first is the distribution strategy.Thu, 06 Aug 2026 00:00:00 GMTmodel-apicompatibilityecosystemdistributionhttps://musecodes.io/blog/multimodal-coding/https://musecodes.io/blog/multimodal-coding/The most quietly radical launch demo wasn’t about speed or parallelism. It was a video file becoming a website.Thu, 06 Aug 2026 00:00:00 GMTmultimodalvisionvideo-inputagentshttps://musecodes.io/blog/kernel-optimization-case-study/https://musecodes.io/blog/kernel-optimization-case-study/Benchmarks measure an agent’s sprint. Meta’s kernel-optimization study measures its marathon — and it’s the more interesting number.Wed, 05 Aug 2026 00:00:00 GMTgpu-kernelslong-horizonagentic-codingtritonhttps://musecodes.io/blog/muse-spark-12-training/https://musecodes.io/blog/muse-spark-12-training/“Built for Muse Spark” isn’t marketing shorthand — it describes a training strategy. Three techniques from Meta’s release notes explain why the model and the agent perform better together than either would with a stranger.Wed, 05 Aug 2026 00:00:00 GMTmuse-sparktrainingco-trainingself-improvement