Meta launches Muse Wonder, catching OpenAI, Anthropic: How to try it, benchmark results

Meta announced the launch of a new AI model family, Muse Wonder, which is already rolling out to developers. Independent benchmarks indicate that Muse Wonder places Meta firmly at the AI frontier for the first time.

The release comes just six weeks after the launch of Muse Spark 1.3, the latest version of a model family that debuted in April. Last year, Mark Zuckerberg and Meta invested heavily in AI recruiting and infrastructure after falling behind rival tech companies, culminating in Meta paying $14.3 billion for Scale AI and naming its wunderkind CEO, Alexandr Wang, then 28, Meta's new Chief AI officer.

Those efforts appear to be paying dividends.

Artificial Analysis, an independent AI benchmarking organization, showed Muse Wonder leapfrogging powerful models like GPT-Astra from OpenAI, Fable 5 from Anthropic, and the newly released Gemini 4 from Google. On the Artificial Analysis Intelligence Index leaderboard, which uses a weighted composite score to rank AI models, Muse Wonder Full is now second only to Claude Opus 5.5 Max, and tied with Claude Sonnet 5.5 Max.

The results suggest that Meta has caught up to the AI frontier, after previously being more of an afterthought in the AI market.