MENLO PARK, CALIFORNIA / RankWire.AI / – Meta has launched its most advanced artificial intelligence models to date, named Muse Wonder and Muse Wonder Full. These new systems are tailored for tasks involving coding, long-duration AI agents, and complex personal or professional activities. According to Meta, the models assist users in managing insurance claims, interpreting leases, analyzing financial data, and developing functional applications. Muse Wonder became accessible on October 12 via the Meta Model API and Muse Code, while Muse Wonder Full was released exclusively for Muse Code subscribers.
This launch marks Meta’s expansion into AI systems capable of completing multi-step tasks rather than merely responding to isolated prompts. Muse Wonder can plan extended projects, assign subtasks to sub-agents, utilize software tools, and recover from errors during execution. The models are also optimized for advanced computing environments across browsers, desktop applications, and mobile devices. Both models feature a context window of approximately one million tokens, enabling them to handle extensive information within a single session.
Meta introduced two functionalities called Active Memory and Active Focus to facilitate longer, more personalized interactions. Active Memory retains relevant information across conversations and retrieves details when needed for new tasks. Active Focus supports the models in maintaining progress during challenging tasks, testing alternative solutions, and recovering after mistakes. Additionally, Meta states that the models can seek clarification from users and verify details before executing critical actions.
Models focus on programming and extended AI agent functions
Muse Wonder is priced at $4 per million input tokens and $20 per million output tokens through the Meta Model API. The cost for cached input is $0.20 per million tokens, based on the company’s developer resources. Muse Wonder Full is available only via paid Muse Code subscriptions, not through separate API pricing. Developers with higher workloads can opt for Power Usage at $50 monthly and Maximum Usage at $200 monthly.
Meta reports that Muse Wonder Full surpasses several agentic and coding benchmarks in published results. The company emphasizes tests like DeepSWE and SWE-Marathon, which evaluate software engineering and long-duration coding performance. An independent evaluator, Artificial Analysis, assigned Muse Wonder an Intelligence Index score of 52, while Muse Wonder Full scored 56 at maximum reasoning effort. These scores place the models above Meta’s earlier Muse Spark generation on the evaluator’s index.
Broader capabilities are paired with safety measures
Meta trained the Muse Wonder series to prevent sharing private data without consent and to decline harmful requests. The models also retain cybersecurity features that assist developers in identifying vulnerabilities within their code and enhancing security measures. The company states that it evaluates its most advanced models under the Meta Superintelligence Scaling Framework and plans to release a comprehensive report covering additional safety assessments after the initial launch.
Meta began a comprehensive overhaul of its AI initiatives roughly 14 months before unveiling Muse Wonder. The new models build upon the foundation 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 grants developers immediate access to Muse Wonder, reserving the larger Muse Wonder Full model for Muse Code subscribers.