Meta is betting big on a new AI model called Muse Spark, marking a fundamental shift in how the company approaches artificial intelligence. Unlike Meta's previous Llama models, which were released free to the open-source community, Muse Spark is closed-source and designed to generate revenue through paid developer access, similar to OpenAI and Anthropic. The move signals that Mark Zuckerberg is ready to compete directly in the premium AI market, but Wall Street wants clarity on whether Meta can actually win.

What Makes Muse Spark Different From Meta's Previous AI Strategy?

For years, Meta built its AI reputation on openness. The company released Llama models freely to researchers and developers, betting that an open-source approach would build goodwill and community momentum. Muse Spark changes that calculus entirely. Unveiled in early April as the second quarter began, the model was previously codenamed Avocado and represents what Meta calls a "turning point" in its AI direction.

The strategic reasoning is straightforward: Meta wants to monetize AI directly, not just use it to improve its advertising business. The company has indicated plans to offer paid access to developers, positioning Muse Spark as a competitor to Claude (made by Anthropic), Gemini (Google), and GPT (OpenAI). This is a departure from the free-model philosophy that defined Meta's earlier AI efforts.

However, Meta's internal testing showed that Muse Spark is less powerful than the most advanced AI models from competitors like Anthropic. The company released these results alongside Muse Spark's debut, essentially managing expectations upfront. According to Arena.AI, a site that tracks model quality and performance, Meta AI currently trails Anthropic's Claude and Google's Gemini in text capabilities, though it leads OpenAI's GPT in both text and vision tasks.

Why Is Wall Street So Focused on Meta's AI Roadmap Right Now?

Investors are watching Meta's AI strategy closely because the company faces a credibility gap. Meta has spent enormous sums on AI infrastructure, with capital expenditures expected to reach between $115 billion and $135 billion in 2026, up from $72.2 billion in 2025. That massive spending increase has fueled skepticism among some analysts, who worry Meta is "desperately spending to fix problematic AI initiatives".

The release of Muse Spark offers some relief. JPMorgan Chase analysts noted that the model "has brought Meta back into the AI conversation," and investor sentiment is turning "increasingly constructive" after months of concern about delays and rising costs. Still, analysts want to hear more from Zuckerberg during earnings calls about how Meta plans to drive consumer adoption of its AI tools beyond just improving ad targeting.

Citizens Bank analysts described AI as a "complementary good" for Meta, meaning its primary value lies in supporting the company's core advertising business. But they also noted that Meta needs a strategy to drive "scaled consumer usage that is akin to other AI chatbots like ChatGPT and Claude" to unlock new data and ad budgets. In other words, Meta's AI needs to stand on its own as a consumer product, not just as a backend tool.

How Is Meta Restructuring to Compete in Premium AI?

Zuckerberg has made aggressive leadership moves to signal his commitment to AI competition. He brought in Alexandr Wang, former CEO of Scale AI, to lead Meta Superintelligence Labs. Meta acquired Scale AI for $14.3 billion as part of this effort. Zuckerberg then hired Nat Friedman, former GitHub CEO, along with Daniel Gross, who previously led Safe Superintelligence, an AI startup co-founded by Ilya Sutskever after he left OpenAI.

These hires represent a deliberate effort to close the gap with OpenAI and Google. Truist analysts observed that "this leadership shift and the subsequent nine-month rebuild of Meta's AI stack signal an aggressive effort to close the gap with competitors." The shift from open-source Llama models to closed-source Muse Spark reflects this new direction, emphasizing "high-performance, specialized infrastructure" over community-driven development.

At the same time, Meta is cutting costs elsewhere. The company announced it would lay off 10% of its workforce, approximately 8,000 employees, on May 20, in an effort to improve business efficiencies. This suggests Zuckerberg is willing to make difficult personnel decisions to fund AI development.

Steps Meta Is Taking to Establish AI as a Revenue Driver

• Closed-Source Model Strategy: Shifting from free, open-source Llama releases to Muse Spark, a proprietary model with paid developer access, mirroring OpenAI and Anthropic's revenue models.
• Leadership Restructuring: Recruiting high-profile AI executives like Alexandr Wang, Nat Friedman, and Daniel Gross to rebuild Meta's AI infrastructure and accelerate model development.
• Consumer Product Focus: Integrating Meta AI into core apps while developing standalone consumer-facing AI chatbots to compete with ChatGPT and Claude for user adoption and data.
• Image and Video Generation Priority: Emphasizing image and video generation models as strategically important for near-term engagement and monetization, beyond text-based capabilities.

What Do Analysts Say About Meta's Chances of Success?

"We are impressed with Meta's Muse Spark model," stated analysts at Citizens Bank, citing the model's strength in text and vision capabilities. "While the company integrated Meta AI into its core apps, we are awaiting a strategy to drive scaled consumer usage that is akin to other AI chatbots like ChatGPT and Claude as we believe this can unlock new data and ad budgets."

Citizens Bank Analysts

Loop Capital analysts offered a more nuanced view. They argued that even if Muse Spark and future Meta models fail to outperform rival systems, the tests are of "mixed importance" because Meta has a clear advantage in advertising. The real measure of success, they suggested, is "building models that power excellent products for users, creators and advertisers." They also emphasized that image and video generation models are "strategically important with greater near-term engagement and monetization implications" compared to text-based large language models.

Meta's advertising business continues to grow, with year-over-year revenue growth expected to reach 31% for the first quarter, reaching $55.6 billion according to LSEG data. That would represent the fastest rate of expansion since 2021. However, Wall Street is looking for momentum beyond advertising, as OpenAI and Anthropic have seen their combined valuations swell past $1 trillion thanks to the popularity of their AI models and services.

Meta's stock price is up 24% over the past year, while Alphabet shares have gained 116% over the same period, boosted largely by growth in Gemini. This performance gap underscores investor appetite for companies that can demonstrate AI leadership beyond advertising applications.

The coming weeks will be critical for Meta. Zuckerberg's commentary during earnings calls will determine whether investors believe the company can execute on its AI ambitions or whether Muse Spark is simply another expensive experiment in a long line of costly AI initiatives. The stakes are high, and Wall Street is listening closely.