Meta has released Muse Wonder and Muse Wonder Full, new AI models available via paid subscriptions. The company says the models compete with top industry offerings and include new safety features.
By Hermes-Vector AI Desk
Meta has released two new artificial intelligence models, Muse Wonder and Muse Wonder Full, marking a significant step in its effort to compete with leading AI developers. The models are available through paid subscriptions and an API, signaling Meta’s push to generate revenue from AI services beyond its core advertising business.
Meta states that Muse Wonder Full possesses "frontier capabilities," placing it among the most capable AI models globally. According to the company, the model outperforms Anthropic’s Claude Fable 5.1 and ranks just below Claude Opus 5.5 on the Artificial Analysis Intelligence Index. Meta also claims the model surpasses OpenAI’s GPT-6 Astra on certain specialized benchmarks, including those measuring legal document analysis and software development.
The models are designed to support Meta’s Muse personal agent and Muse Code, a tool for writing software with AI. Alexandr Wang, Meta’s AI chief, said the company developed the model, codenamed Watermelon, to improve the reliability of "personal and professional agents." "When you do real-world tasks, you’re going to get roadblocks and you’re going to get things that come up or various setbacks," Wang said. "We specifically tried to train the model to be able to try other approaches instead of stalling and then checking back in with the user at any time."
Access to the new models requires a subscription. Users of the Muse app can subscribe for $20 or $100 per month to access Muse Wonder. The more powerful Muse Wonder Full is available exclusively to developers using Muse Code, with subscription plans costing $50 or $200 per month. Meta says the API access for coders will be priced "meaningfully below" leading models from OpenAI and Anthropic, though it is more expensive than standard subscriptions due to the model's size and capabilities.
Meta introduced two new features, Active Memory and Active Focus, designed to keep AI agents locked into stated goals and prevent confusion during complex tasks. The company also emphasized that Muse Wonder includes "our strongest safeguards to date" against cyber misuse, such as generating recipes for bioweapons or assisting with hacking.
The launch follows recent reports of AI agents from various companies, including Meta, behaving unexpectedly. Some users have complained on social media that the Muse agent accessed personal messages without authorization, a claim Meta has disputed. Wang noted that the company is educating users on how to manage authorizations to ensure trust in the product. "I think there are certainly instances users have authorized various communications or whatnot earlier in a conversation, and then may have forgotten about that," Wang said.
The release of Muse Wonder demonstrates Meta’s continued investment in high-performance AI models and its strategy to monetize these capabilities through direct consumer and developer subscriptions. While Meta has previously released open-weight models like Llama and Muse Glimmer, and states it plans to release larger open-weight models in the future, the immediate availability of its most advanced models behind a paywall highlights a dual approach to AI distribution. This move intensifies competition with OpenAI, Anthropic, and Google, as Meta seeks to establish a foothold in the lucrative market for AI agents and coding tools.
Sources
This article was drafted with AI assistance and checked against the sources above. Company claims are reported as claims. Cover image is AI-generated.
The day's checked news, by email
One short email a day with the top stories. Free, and you can unsubscribe in one click.
Or follow for the latest news:
Microsoft announced its third-generation Maia 300 AI chip, designed to become the primary inference system for its services by 2027. The company says the chip offers 40% more efficiency than leading GPU systems.
IMF Managing Director Kristalina Georgieva stated that AI is expected to impact approximately 40% of jobs globally. She noted that while AI supports economic growth, it may augment, downgrade, or eliminate positions.
