The new suite features a 1 million token context window and pricing of $4 per million input tokens and $20 per million output tokens for the standard Muse Wonder model, which undercuts rivals such as Anthropic. On the Artificial Analysis Intelligence Index, Muse Wonder scored 56, making it the second best model globally and just 2 points behind the current leader, Opus 5.5. The release is the result of a year of development by Meta Superintelligence Labs.
Tuesday, October 20, 2026
Story
Meta’s Muse Wonder AI Beats Rivals on Agent Coding Benchmarks
AI8 days ago
Meta has introduced two new large-scale AI models designed to automate professional tasks and handle long-horizon reasoning. These frontier models, branded Muse Wonder and Muse Wonder Full, target personal and professional agentic work and are now available through the company's API and Muse Code. Muse Wonder Full reported a 57.5% score on the SWE-Marathon benchmark, exceeding the performance of Claude Opus 5.5 at 46.3% and GPT-6 Astra at 42.5%.
Looking for more context or analysis?
Tilt chat provides reliable, tailored insights, alerts, and automations.