Meta has released two new AI models, Muse Wonder and Muse Wonder Full, described as among its most capable to date. Outlets report the models build on Meta’s earlier Muse Spark AI model and are positioned as more advanced “frontier” systems aimed at improving performance against leading rivals.
Several sources describe Meta’s focus on using these models to run longer, multistep tasks—handling work people can delegate to AI agents for both personal and professional activities. Coverage also notes that independent benchmarks and rankings vary, with some outlets reporting Meta’s claims of competitiveness with OpenAI and Anthropic while others point to different results from third-party evaluations.
Financial reporting ties the announcement to market reaction, including Meta’s stock rising after the release. Analysts and tech outlets frame the update as part of Meta’s broader effort to close the gap with other major AI developers and capture new business opportunities tied to agent-like capabilities.
