MENLO PARK, CALIFORNIA / RankWire.AI / – Meta has unveiled the Muse Wonder and Muse Wonder Full models, a new lineup of sophisticated artificial intelligence systems. These models are designed for tasks involving software development, AI agents, and complex daily activities. They can support functions such as managing leases, processing insurance claims, handling financial data, and developing applications. Meta launched Muse Wonder via its Model API and Muse Code on October 12, with Muse Wonder Full becoming accessible to users holding eligible Muse Code subscriptions.
Focusing on multi-step processes rather than single responses, the models are capable of planning workflows, utilizing digital tools, and breaking down large jobs into smaller, specialized sub-agents. They function across browsers, desktop programs, and mobile platforms. Meta engineered these systems to recover gracefully when an approach fails. Supporting context windows of approximately one million tokens, they can analyze extensive documents, codebases, and other substantial content within a single session.
Additionally, Meta introduced features called Active Memory and Active Focus alongside the Wonder models. Active Memory retains relevant information from previous conversations and retrieves it as needed. Active Focus enables an AI agent to persevere through difficult tasks and adapt its strategy after encountering errors. The models can ask for clarification when instructions lack crucial details and seek confirmation before executing sensitive or critical actions. These enhancements are intended to facilitate longer work periods without forcing users to restart tasks repeatedly.
Meta’s new models boost agent and coding capabilities
Pricing for Muse Wonder is set at $4 per million input tokens via the Meta Model API. Output costs are $20 per million tokens, while cached input is priced at $0.20 per million tokens. Instead of separate API fees, Meta offers Muse Wonder Full through paid Muse Code plans. The company also provides a Power Usage plan at $50 monthly and a Maximum Usage plan at $200 per month, targeted at developers with higher demands for intensive coding tasks.
Meta shared benchmark data related to software engineering, coding agents, and extended technical operations. The tests included DeepSWE and SWE-Marathon, with Artificial Analysis evaluating both models at maximum reasoning effort. Muse Wonder received an Intelligence Index score of 52, while Muse Wonder Full scored 56. These results surpass Meta’s previous Muse Spark models on the evaluator’s scale, indicating a clear improvement in reasoning performance based on testing metrics.
Safety and privacy measures accompany the new release
Meta emphasizes that the Wonder models incorporate safeguards to protect privacy, prevent harmful requests, and ensure cybersecurity. They are trained to avoid revealing private information without user consent and to decline requests involving harmful activities, all while maintaining essential security functions. Developers can utilize these models to identify vulnerabilities in their own software and bolster defenses. Meta is evaluating its most advanced AI systems under its Superintelligence Scaling Framework and plans to release additional details about safety testing for the new models.
The rollout follows a restructuring of Meta’s AI initiatives that began roughly 14 months ago. The company states that Muse Wonder builds upon technical foundations laid by Muse Spark. Development of the new models took place at Meta’s Prometheus data center. The firm also intends to integrate Muse Wonder technology into more features within the Muse platform, starting with subscribers. While developers can already access Muse Wonder, Muse Wonder Full remains available through qualifying Muse Code subscription plans.