Late on the night of July 9, Mark Zuckerberg reactivated his long-dormant X account @finkd to unveil Meta's latest model, Muse Spark 1.1, which immediately claimed the top spot across three major professional benchmarks: tax, healthcare, and legal, dethroning Grok 4.5. The model is priced at just one-tenth the cost of Fable 5, with Zuckerberg describing it as "very low cost." Muse Spark 1.1 is the second-generation multimodal reasoning model from Meta's Superintelligence Lab, featuring a 1 million token context window that can self-manage and self-compress. It functions as an Agent, breaking down complex tasks, dispatching sub-Agents, and making autonomous judgments for computer operations. In coding scenarios, it handles debugging, new feature development, and large-scale code migrations with native compatibility for OpenCode, Cline, and Replit. Pricing is set at $1.25 per million input tokens, and $4.25 per million output tokens, making it 8 times cheaper for inputs and nearly 12 times cheaper for outputs compared to Anthropic's Fable 5.