Meta Platforms (META) Stock Analysis 2026: AI Spending Hits $145B, Llama 4 Lands — Is the Bull Case Still Intact?
July 3, 2026 · 10 min read
Meta's Q1 2026 earnings were a study in contradictions — revenue surged 35% to $57.3B, beating estimates, while the stock dropped 8.9% as Zuckerberg raised the 2026 capex plan to $145B. AI glasses are tripling their daily user base. Llama 4 is setting open-source benchmarks. Yet investors are asking: when does all this spending pay off?
META At-a-Glance — July 2026
Current Price
~$580
As of early July 2026
Market Cap
~$1.5T
5th largest globally
Q1 2026 Revenue
$57.3B
+35% YoY; beat estimates
Q2 2026E Revenue
$58–61B
Company guidance
2026 CapEx Guidance
$125–145B
Up from $115–135B
AI Glasses DAU
3× YoY
Ray-Ban Meta glasses
Fwd P/E
~26x
Cheaper than MSFT/NVDA
Consensus Target
$640
+10% from current
Quarterly Revenue — FY2024 to FY2026
Meta's revenue trajectory has been remarkably consistent since the 2022 "Year of Efficiency" cost-cutting that restored margins. Revenue has grown every quarter since Q4 2022, with the AI-enhanced advertising engine compounding growth despite a maturing social media market.
Q1'24
$36.5B
Q2'24
$39.1B
Q3'24
$40.6B
Q4'24
$48.4B
Q1'25
$42.3B
Q2'25
$45.2B
Q3'25
$47.3B
Q4'25
$55.6B
Q1'26
$57.3B
Q2'26E
$59.5B
What Actually Drives Meta's Revenue
Despite all the AI and metaverse headlines, Meta is still — overwhelmingly — an advertising business. Understanding the ad flywheel is the key to understanding the stock.
Family of Apps (FB, Instagram, WhatsApp, Threads)~$52B/qtr (98%)
DAU growth (Instagram + FB)PositiveBoth platforms growing globally
AI Glasses DAU growth (YoY)3× tripledRay-Ban Meta smart glasses
Forward P/E~26xDiscount to MSFT (32x), NVDA (42x)
DividendNoneReturns capital via buybacks
The $145B Question: What Is Meta Buying?
Meta's decision to raise capex to $145B for 2026 is the central investor debate. Here is exactly where the money is going and why Zuckerberg believes it is necessary.
Training next-generation Llama models
Training Llama 4 and the future Llama 5 requires clusters of 100,000+ AI accelerators running for months. Meta is building proprietary AI training infrastructure — including custom MTIA AI chips — to reduce dependence on Nvidia and lower training costs over time. This compute investment is not optional; it is the price of staying competitive with OpenAI, Google, and Anthropic.
Data center construction across three continents
Meta is building what it calls the largest AI training cluster in the world — a 2 gigawatt facility that will house millions of AI accelerators. Additional data centers are under construction in Europe (to serve GDPR-compliant AI) and Asia. The physical construction cost — land, buildings, power infrastructure — accounts for roughly 40% of total capex.
AR/VR hardware and optics R&D
Reality Labs continues to absorb $4B+ per year in operating losses, funding the development of next-generation AR glasses (Orion prototype), the Quest 4 headset, and the underlying display and optics technology. Zuckerberg believes AR glasses will be as significant as smartphones by 2030. That thesis is expensive to pursue, but Ray-Ban Meta glasses tripling daily users is an early validation signal.
Inference infrastructure for 3B+ users
Meta AI (the assistant) is embedded in WhatsApp (3B users), Instagram (2B+ users), Facebook, and Messenger. Serving AI responses to this user base at low latency 24/7 requires a different kind of infrastructure than training — inference servers optimised for throughput and availability. Meta is building dedicated inference capacity globally.
Meta's AI Products: What's Working in 2026
Meta AI Assistant
Embedded across WhatsApp, Instagram, and Facebook, Meta AI is available to 3B+ monthly active users. While monetisation is still limited, daily engagement with the assistant is growing and building the habit loop that Meta will eventually monetise via premium features or commerce integrations.
Llama 4 (Open Source)
Llama 4 is competitive with proprietary models from OpenAI and Anthropic on standard benchmarks. By releasing it openly, Meta commoditises AI models, undermines rivals' pricing power, and positions itself as the AI provider for small businesses — who then use Meta's advertising tools.
Advantage+ Advertising
AI-automated campaign management that adjusts bids, audiences, and creative in real time. Advertisers using Advantage+ see an average of 22% lower cost per result versus manual campaigns. This makes Meta's ads demonstrably more effective, driving higher CPMs and ARPU across the Family of Apps.
Ray-Ban Meta Glasses
Daily active users of Ray-Ban Meta smart glasses tripled YoY, making it the first consumer AR product with genuine traction. While still a niche device ($299–399), the glasses represent Meta's on-ramp to AR wearables ahead of a full AR display launch (Orion) expected in 2027.
Bull Case: Why META Could Reach $700+
✓
AI advertising compounds ARPU for years: Meta's revenue per user (ARPU) in North America is already the highest in the industry. AI-optimised advertising is systematically raising CPMs globally, particularly in markets where ARPU was historically low (Southeast Asia, Latin America, India). If Advantage+ and AI creative tools keep improving, Meta can grow revenue without growing users.
✓
Meta AI becomes a monetisable assistant: With 3B+ monthly active users across its apps, Meta has a larger AI assistant distribution than Google Assistant, Siri, or Alexa. A paid Meta AI subscription at $10–15/month would generate $30–50B in new annual revenue with high margins. Even 5% conversion of WhatsApp users alone would be transformative.
✓
The capex resolves into a durable moat: Microsoft, Google, and Amazon have been building AI infrastructure for years and have the systems to show for it. When Meta's $145B capex is fully deployed by 2027, it will have infrastructure rivalling the hyperscalers — at a fraction of the enterprise cloud cost, because it serves its own workloads. The stock will re-rate when the market sees FCF recovering.
✓
AR glasses become the next platform: If Ray-Ban Meta glasses are the early iPhone moment and Orion (true AR display, expected 2027) delivers, Meta is positioned to own the wearable computing category. The AR addressable market — spatial computing, commerce, communication — is estimated at $1T+ over the decade.
Bear Case: Why META Could Disappoint
✗
Capex discipline gone — ROI unclear: At $145B in 2026 alone, Meta is spending at a rate no private company has ever sustained. If AI monetisation does not materialise on schedule, shareholders are funding a multi-year science project. The history of massive tech capex cycles (Microsoft's phone acquisition, Google's hardware push) includes many expensive failures.
✗
Digital advertising market slows: Meta's entire business depends on digital ad spending. A US recession, a significant TikTok resurgence, or a platform shift toward video games or other non-ad media would directly compress revenue. The advertising market is cyclical, and Meta's high growth rates depend on an environment where businesses continue increasing digital marketing budgets.
✗
Regulatory pressure intensifies globally: The EU's Digital Markets Act designates Meta as a 'gatekeeper' and imposes strict data and interoperability requirements. Regulatory actions in the US (FTC antitrust), UK (CMA), and India add compliance costs and cap certain growth strategies (targeted advertising, data sharing across apps).
✗
Reality Labs is an expensive distraction: Reality Labs has lost more than $50B cumulatively since being disclosed as a segment in 2021. If the AR glasses pivot doesn't gain mass market traction by 2028, shareholders may demand Zuckerberg shut it down. The persistent losses reduce total operating income and distract management attention from the core advertising business.
Analyst Ratings & Price Targets — July 2026
Oppenheimer
$720
Buy
Bank of America
$700
Buy
JPMorgan
$680
Buy
Citi
$640
Buy
New Street
$550
Hold
Oppenheimer
Buy
$720
Ad market share gains + AI-driven ARPU expansion
Bank of America
Buy
$700
Llama 4 cuts infra costs; advertising flywheel intact
JPMorgan
Buy
$680
Instagram Reels + AI personalisation driving engagement
Citi
Buy
$640
Capex overhang fades as investors see ROI signals
New Street
Hold
$550
$145B capex with unclear AI monetisation path
Consensus target: ~$640 · Implied upside from $580: +10%
Bottom Line: Is META a Buy at $580?
Meta is in the paradox of being one of the world's most profitable companies while voluntarily compressing its own free cash flow through a $145B capex bet on AI. The advertising engine is working — 35% revenue growth from a $57B base is exceptional. The question is whether the capital allocation will look visionary (like AWS from 2006–2016) or wasteful (like Microsoft's Nokia acquisition).
At 26x forward earnings, META is the cheapest of the Magnificent Seven AI plays. The discount reflects capex uncertainty, not fundamental weakness. Investors with a 3-year view are essentially betting that the $145B in AI infrastructure will generate $200B+ in incremental revenue by 2028–2029. Historically, Zuckerberg's large bets — Instagram, WhatsApp, Reels — have paid off with remarkable consistency.
Our framework (not financial advice)
Long-term AI believer?META at 26x is arguably the best-value AI mega-cap; the capex story is already known and partially priced
Concerned about capex?Wait for Q2 2026 earnings (late July); look for management commentary on capex discipline or AI monetisation proof
Already in a MAG-7 position?META rounds out AI exposure with a social/ad angle distinct from cloud (MSFT, GOOGL) and hardware (NVDA)