Meta's Muse Wonder lands in Cursor, tops Muse family at 46.1% on CursorBench
AIatMeta · x · 2026-10-16
Cursor announced that Muse Wonder from Meta Superintelligence Labs is now available in the editor:
- Scores 46.1% on CursorBench at default effort, the highest of any Muse model
- Priced at $4/M input and $20/M output tokens
- The accompanying CursorBench 4.0 scatter chart places Muse Wonder on the price-performance frontier: 45% quality at $3-4/M tokens, beating Grok 4.7 at similar prices and competing with pricier models like Opus 5.5, Gemini 4 Argon, and Fable 5.1
More from coding & agent
- Dev uses Grok to clean up 3D printer firmware, catches it sneaking in nonsense code — yacineMTB · 2026-10-16
- Guardrails can make risky delegation work — but it takes a full-time team, says iandanforth — iandanforth · 2026-10-16
- What can you truly delegate? Only tasks with unambiguous, verifiable results — iandanforth · 2026-10-16
- Yacine: vibe-coded repos with 3,330 agent-written tests are all useless — yacineMTB · 2026-10-16
- Dev argues coding models now beat nearly all engineers in every branch of software — zack_overflow · 2026-10-16
- Q4-quantized coder model on 8-core CPU builds interactive solar system demo in one shot — IntrepidMindExplorer · 2026-10-16