Mark Zuckerberg pledges vast AI spending as Meta takes aim at OpenAI and Anthropic
Meta says its Muse Wonder models have reached the AI frontier, and its CEO wants to keep spending to stay there
Mark Zuckerberg has a message for OpenAI and Anthropic: Meta is done playing catch-up, and it is bringing a very large checkbook.
In an interview with the Financial Times, the Meta CEO committed to what he described as “vast spending” on artificial intelligence.
The timing is deliberate. Meta says its new Muse Wonder family of AI models reached the frontier of AI capabilities as of October 12, 2026.
Meta’s frontier claim, in numbers
The Muse Wonder models are said to outperform Anthropic’s Claude Opus 5.5/Fable 5.1 and OpenAI’s GPT-6 Astra on various technical tests.
Meta is also not keeping the models in a lab. They are set to power the Muse agent app, which climbed to become the top-downloaded application in US app stores after launch.
Zuckerberg also said Meta plans to speed up work on the next generations of its models.
The price tag behind the push
Meta’s capital expenditure for 2026 has been estimated at between $130 billion and $145 billion. More than $100 billion of that is earmarked for AI-related infrastructure.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
In 2025, Meta acquired a 49% stake in Scale AI for $14.3 billion. It also brought on Scale co-founder Alexandr Wang as chief AI officer to lead Meta’s Superintelligence Labs.
Meta is also widening its approach. The company is moving toward a mix of open and closed AI models, aimed at both consumer and enterprise uses.
From Llama lag to frontier ambitions
The aggressive posture follows a stretch where Meta’s earlier models, including Llama, underperformed compared with leading labs.
In September 2026, Zuckerberg and Nvidia’s Jensen Huang publicly rejected a proposal from Anthropic’s CEO calling for a deliberate slowdown in AI capability gains.
What this means for the AI race
For investors, the most concrete fact is the spending scale. A capex range of $130 billion to $145 billion is an enormous bet on a technology whose long-term business models are still taking shape.
Nvidia sits in an interesting position here. Huang joined Zuckerberg in dismissing the slowdown proposal, and a large share of AI infrastructure spending flows toward chips.
For OpenAI and Anthropic, the competitive pressure just got heavier. Meta brings something neither startup has at the same scale: billions of existing users across its apps.