Meta released Muse Wonder Full on October 12, 2026, saying it closes the gap to leading frontier models. It is the company's first major update to its AI models since the release of Muse Spark 1.3.
Meta reported Muse Wonder Full scores 56 on the Artificial Analysis Intelligence Index, measured on v4.3, which compares with 48 for its predecessor, Muse Spark 1.3.
Muse Wonder Full is built on a larger base model and targets complex tasks like coding and problem-solving. Availability begins with Muse Code users, initially for developers and advanced users.
"Active Memory lets the model remember key information across conversations and recall it when needed," said Matthias Bastian, Meta's AI strategist. The feature is central to Meta's pitch for a personal AI agent that can access calendars, messages, and documents.
The announcement follows a weaker stretch with Meta's open Llama models. Meta says it has recovered, with Muse Wonder Full placing in the top three on six out of ten benchmarks overall.
Meta did not say how the model handles real-world tasks beyond training data, and it raised concerns about factual knowledge lagging behind competitors. The company said it plans to build specialized models from Muse Wonder Full over time.
Source: thedecoder
