Meta Releases Muse Wonder AI Models to Rival OpenAI and Anthropic
Meta has released two flagship AI models, Muse Wonder and Muse Wonder Full, designed to compete with top industry leaders in reasoning, writing, and coding. These models were developed under the leadership of Chief AI Officer Alexandr Wang and will power features in Meta's consumer and developer applications. This release marks Meta's first direct attempt to challenge top-tier frontier models from OpenAI and Anthropic, signaling an aggressive push into enterprise AI services and advanced agentic workflows. By integrating these models into products used by billions, Meta is expanding its competitive reach in the generative AI market. Muse Wonder Full is a more powerful version restricted to subscribers of Meta's Muse Code programming tool, while both models were trained at Meta's gigawatt-scale Prometheus data center. The development faced a temporary pretraining pause in July before launching this October.
## BACKGROUND
AI models with advanced reasoning, coding, and writing capabilities are increasingly used to build autonomous software agents that can execute complex tasks across various applications. Large tech companies are heavily investing in these frontier models to capture both consumer markets and lucrative enterprise software development sectors.