Meta unveils Muse Spark AI model to rival top chatbots

Watch on YouTube ↗  |  April 08, 2026 at 18:36  |  4:08  |  CNBC

Summary

  • Meta unveils its new AI model "Muse Spark," its most powerful model yet, launching in the Meta AI app and website.
  • The model represents a more limited initial release than expected, focusing on custom uses for Meta's apps like shopping and visual coding tools.
  • It is positioned as the start of a rollout, with more sophisticated "frontier" models (competitive with Gemini, Deep Think, GPT-3 Pro) planned for later.
  • Meta plans to eventually offer paid API access for the model, opening a potential new AI revenue stream.
  • Steve Weiss views the launch as an early return on Meta's massive AI investment, strengthening its moat and creating new products for advertising and subscription fees.
  • Weiss believes Meta's transformation is occurring in real-time and will continue, but notes the stock is not inexpensive after the day's 6.5% rally.
  • Broader large-cap tech (Amazon, Alphabet) also rallied, with sentiment driven by strong anticipated tech earnings growth (45% YoY cited) and investor appetite.
  • The market reaction suggests investors are eager to position ahead of earnings season for these major tech names.
Trade Ideas
Steve Weiss Chief Investment Officer, Short Hills Capital Partners 2:37
Steve Weiss states he owns META, it is now his largest position, and he would "definitely add to it." He sees the Muse Spark launch as an early return on Meta's major AI investment. The AI investment is seen as strengthening the company's competitive moat and creating new products for advertising and subscription fees, driving a transformation. LONG due to the ongoing transformation and the expectation of continued returns from AI investments, despite acknowledging the stock is not cheap. Valuation; the stock may be fully priced after the rally, limiting near-term upside.
Up Next

This CNBC video, published April 08, 2026, features Steve Weiss discussing META. 1 trade idea extracted by AI with direction and confidence scoring.

Speakers: Steve Weiss  · Tickers: META