{"slug": "cursor-adds-metas-muse-wonder-model-which-scores-46-1-on-cursorbench", "title": "Cursor adds Meta’s Muse Wonder model, which scores 46.1% on CursorBench", "summary": "Meta Superintelligence Labs released Muse Wonder on October 12, 2026, and the agentic coding model is now available inside Cursor, where it scored 46.1% on CursorBench 4.0 in its 'High' configuration and up to 50.4% at its 'Max' setting. That leaves Muse Wonder 11 points behind the benchmark leader and 7.4 points behind Claude Opus 5.5's 57.8% maximum-effort result, with the model priced at $4 per million input tokens and $20 per million output tokens through the Meta Model API. Muse Wonder ships with a 1M-token context window and accepts text, images, video and audio, and Meta says its internal evaluations show strengths in long-horizon coding and agent workflows.", "body_md": "Photo: Steve A Johnson / Pexels\n\n# Cursor adds Meta’s Muse Wonder model, which scores 46.1% on CursorBench\n\nMeta Superintelligence Labs' new agentic model lands in the popular AI code editor with a mid-pack benchmark score and a 1M-token context window\n\n[Meta](https://cryptobriefing.com/markets/meta/) Superintelligence Labs has a new model in one of the most competitive rooms in AI: the code editor. Muse Wonder is now available inside Cursor, and it posted a score of **46.1%** on CursorBench.\n\nThat is a respectable debut, not a coronation. The current leader on the same benchmark sits more than 11 points higher, so Meta’s entry joins the race somewhere in the middle of the pack.\n\n## What Muse Wonder brings to Cursor\n\nCursor is the AI-powered code editor built by Anysphere. It lets developers choose which underlying model does the heavy lifting, which makes it a kind of open audition stage for AI labs.\n\nMeta Superintelligence Labs released Muse Wonder on October 12, 2026, alongside a more capable sibling called Muse Wonder Full. Both are pitched at agentic work, meaning tasks where the model has to keep a goal in mind across many steps instead of answering one question and stopping.\n\nTo support that, Muse Wonder ships with a 1M-token context window. Tokens are the chunks of text a model reads, so a window that large lets it hold a sizable codebase or document set in working memory at once.\n\nThe model also accepts multimodal input, including text, images, video and audio.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\nPricing through the Meta Model API is set at **$4 per million input tokens** and **$20 per million output tokens**.\n\n## How the benchmark numbers stack up\n\nThe 46.1% figure comes from CursorBench 4.0, using the model’s ‘High’ configuration. Turned up to its ‘Max’ setting, Muse Wonder can reach as much as 50.4%.\n\nFor comparison, Claude Opus 5.5 leads at 57.8% in its maximum effort mode. Even at full throttle, Meta’s model trails that mark by 7.4 points.\n\nThe research summary describes the result as competitive among mid-tier models and favorable against several other frontier systems.\n\nCursorBench itself is fairly new. Version 4.0 was introduced on September 10, 2026, roughly a month before Muse Wonder launched, and it focuses on long-horizon coding tasks drawn from real-world software development scenarios.\n\nMeta’s own internal evaluations point to strengths in long-horizon coding and agent workflows.\n\n## Where this fits in Meta’s AI push\n\nThe Muse Wonder models are built on Meta’s earlier Muse [Spark](https://cryptobriefing.com/markets/spark-2/) series. They arrive as part of a broader effort to sharpen the company’s AI offerings following a significant restructuring of its AI division in mid-2025.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our\n\n[Editorial Policy](https://cryptobriefing.com/editorial-policy/).", "url": "https://wpnews.pro/news/cursor-adds-metas-muse-wonder-model-which-scores-46-1-on-cursorbench", "canonical_source": "https://cryptobriefing.com/cursor-adds-meta-muse-wonder-cursorbench/", "published_at": "2026-10-15 23:02:04+00:00", "updated_at": "2026-10-15 23:29:36.734055+00:00", "lang": "en", "topics": ["large-language-models", "ai-products", "ai-agents", "developer-tools", "artificial-intelligence"], "entities": ["Meta Superintelligence Labs", "Muse Wonder", "Cursor", "Anysphere", "CursorBench", "Claude Opus 5.5", "Meta Model API", "Muse Wonder Full"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/cursor-adds-metas-muse-wonder-model-which-scores-46-1-on-cursorbench", "markdown": "https://wpnews.pro/news/cursor-adds-metas-muse-wonder-model-which-scores-46-1-on-cursorbench.md", "text": "https://wpnews.pro/news/cursor-adds-metas-muse-wonder-model-which-scores-46-1-on-cursorbench.txt", "jsonld": "https://wpnews.pro/news/cursor-adds-metas-muse-wonder-model-which-scores-46-1-on-cursorbench.jsonld"}}