Meta Muse Spark 1.1: Claude Opus 4.8-level AI model

Meta has announced its new AI model, Muse Spark 1.1, which achieves benchmark scores equivalent to Claude Opus 4.8. This multimodal inference model is a major upgrade. Here's the thing: Muse Spark 1.1 excels at personal tasks across different apps and services. It can manage a one million token context window, remembering past operations through long sessions. What's interesting is its coding performance has significantly improved, especially with large and complex codebases. The model outperformed its predecessor, Muse Spark, and GPT-5.5 in Meta's internal coding evaluation tool, showing comparable performance to Claude Opus 4.8. It even surpassed Claude Opus 4.8 in multiple tests. The bottom line: Muse Spark 1.1 is already being used daily by developers and researchers at Meta and is available in the Meta AI app and for development via the Meta Model API, offering advanced capabilities for various applications.