Meta Launches Muse Spark From Superintelligence Lab
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Meta released Muse Spark on April 8, 2026, the first model from its new Meta Superintelligence Labs led by Alexandr Wang, designed to power "a smarter and faster" Meta AI across the company's products (per FT).
- The model is purpose-built for social media apps, per the Financial Times, with new features rolling out on Facebook and Instagram (per iPhone in Canada) and reasoning capabilities coming to the Meta AI app (per Engadget).
- TechCrunch frames the launch as a "ground-up overhaul" of Meta's AI, while VentureBeat's headline reads "Goodbye Llama" — signaling a strategic shift from Meta's open-source Llama lineage to a proprietary, closed-source model (per Wired).
- Investors are publicly questioning Meta's huge AI investment even as the model debuts (per FT), and the Wall Street Journal frames the launch as a "major test of AI ambitions."
- Decrypt's headline benchmarks Muse Spark as Meta's "most capable" model but notes Google's Gemini Pro "still leads," underscoring how far Meta still trails on raw benchmarks.
- Coverage spans nearly 30 outlets — from Times of India to AFP, WSJ, and Wired — with @aiatmeta, @alexandr_wang, and Scale AI posting launch announcements on X and Threads.
Why it matters: This is the first tangible output of Meta's superintelligence unit — the team Zuckerberg paid top-of-market to build — and the FT reports investors are already publicly questioning whether the billions will pay off. The "Goodbye Llama" framing (per VentureBeat) and the Wired open-vs-closed-source angle mark a real strategic pivot: Meta is abandoning its open-source differentiator just as WSJ calls the launch a "major test of AI ambitions."
Ask SkimNews
