MENLO PARK, CALIFORNIA / RankWire.AI / – Meta has launched Muse Wonder and Muse Wonder Full, its most advanced artificial intelligence models to date. These new systems are tailored for tasks such as coding, managing long-duration AI agents, and handling complex personal or professional responsibilities. According to Meta, the models assist users with activities like processing insurance claims, interpreting leases, analyzing financial data, and developing functional applications. Muse Wonder was made available on October 12 via the Meta Model API and Muse Code, while Muse Wonder Full was introduced for Muse Code subscribers.
This release marks Meta’s expansion into AI systems capable of executing multi-step tasks, moving beyond simple prompt responses. Muse Wonder can organize extended projects, assign subtasks to sub-agents, utilize software tools, and recover from errors during operation. Meta has also optimized these models for use across web browsers, desktop applications, and mobile platforms. Both models support a context window of approximately one million tokens, enabling them to handle vast amounts of information within a single session.
To facilitate longer and more personalized interactions, Meta added features named Active Memory and Active Focus. Active Memory maintains relevant information throughout conversations and retrieves details when they are pertinent to a new task. Active Focus assists the models in persevering through challenging tasks, testing alternative strategies, and recovering after mistakes. Meta also states that the models can seek clarification from users and verify actions before execution.
Models focus on coding and persistent AI agents
The cost of Muse Wonder is $4 per million input tokens and $20 per million output tokens via the Meta Model API. Cached input tokens are priced at $0.20 per million, as per the company’s developer details. Muse Wonder Full is available exclusively through paid Muse Code subscriptions, not separate API pricing. Developers with heavier coding workloads can opt for Power Usage at $50 a month and Maximum Usage at $200 a month.
Meta reports that Muse Wonder Full outperforms several benchmarks related to agentic and coding capabilities. The company highlights tests such as DeepSWE and SWE-Marathon, which evaluate software engineering and long-duration coding performance. Independent evaluator Artificial Analysis awarded Muse Wonder an Intelligence Index score of 52, while Muse Wonder Full achieved a score of 56 under maximum reasoning effort. These scores place the models above Meta’s earlier Muse Spark generation on the evaluator’s index.
Enhanced safety features accompany expanded functionalities
Meta has trained the Muse Wonder series to refrain from sharing private information without permission and to decline harmful requests. The models also retain cybersecurity functions that assist developers in identifying vulnerabilities in their code and strengthening security defenses. The company states that it assesses its most capable models under the Meta Superintelligence Scaling Framework and plans to publish a comprehensive report on additional safety evaluations following the initial release.
Meta initiated a significant overhaul of its AI efforts roughly 14 months before launching Muse Wonder. The company claims the new models are built upon the foundations laid by Muse Spark, with training conducted at its Prometheus data center. The company intends to integrate Muse Wonder into more features within its Muse platform, starting with subscribers. The October release provides developers immediate access to Muse Wonder, reserving the larger Muse Wonder Full model for Muse Code subscribers.