Meta is back on the AI leaderboard with its new family of Wonder models

Meta's Superintelligence Lab released two new frontier models on Monday, Muse Wonder and Muse Wonder Full. These larger models are designed to be more capable than the lightweight Muse Spark, offering higher benchmark scores and improved performance on complex tasks such as coding. Meta's testing indicates the models achieve benchmark scores comparable to GPT-6 Astra and Fable 5.1. The models feature a property called "active memory," which emphasizes the ability to retain and synthesize user information.

According to TechCrunch, Muse Wonder is available via the Meta Model API at a cost of $4.00 per million input tokens and $20.00 per million output tokens, a price point roughly equivalent to Opus 5.5 or GPT-6 Sol. Muse Wonder is also available to Muse app subscribers and Muse Code users. Muse Wonder Full will be available only to Muse Code subscribers and will not be served through the API. Both models will power the Muse agent.

The launch follows a year in which CEO Mark Zuckerberg pledged billions of dollars and recruited researchers to compete in the AI industry, according to Economic Times Tech. The models were developed over 14 months under a revamped AI unit, according to Quartz. MarketWatch Top reports that Meta's stock rose more than 3% following the release.

Brief written by urgent.news from MarketWatch Top, Economic Times Tech, TechCrunch, Techmeme, Quartz — 5 reports on this story. Machine-written — may contain errors; check the original before relying on it.