Release Date: July 29, 2026
For the complete transcript of the earnings call, please refer to the full earnings call transcript.
Positive Points
- Meta Platforms Inc META reported strong Q2 2026 revenue of $60.8 billion, up 28% year-over-year, driven by a 27% increase in ad revenue and a 73% surge in Family of Apps Other revenue.
- AI investments are significantly improving core business performance, with LLMs enhancing content recommendations and ad targeting, leading to an 8.3% increase in ad clicks and a 15.7% uplift in conversions on Facebook.
- The company is rapidly scaling new AI-driven products, including Meta Business Agents used by over 1 million businesses weekly, and Meta One subscriptions, creating new revenue streams beyond advertising.
- Meta's AI models, such as Muse Spark 1.1 and Muse Image, are advancing rapidly, with Muse Spark showing strong agentic capabilities and being distributed via a new public API, positioning the company for enterprise growth.
- Community engagement reached new highs, with 3.6 billion daily active users across apps, Instagram hitting 2 billion daily actives, and Threads surpassing 500 million monthly actives, providing a massive platform for AI-driven innovations.
Negative Points
- Total expenses surged 55% year-over-year to $42 billion, driven by higher employee compensation, infrastructure costs, legal charges, and severance expenses, pressuring operating income.
- GAAP operating income declined 8% year-over-year to $18.8 billion, with operating margin at 31%, reflecting the impact of rising costs despite revenue growth.
- Capital expenditures soared to $31.1 billion in Q2, with full-year 2026 CapEx guidance raised to $130-$145 billion, signaling aggressive infrastructure spending that may weigh on near-term free cash flow.
- Free cash flow dropped sharply to $784 million, down from prior levels, as heavy investments in AI infrastructure and data centers strain liquidity.
- The company faces ongoing legal and regulatory risks, including youth-related trials in the US and other markets, which could result in material losses and further financial uncertainty.
Q & A Highlights
Here are the key highlights from Meta Platforms Inc (META) Q2 2026 earnings call, focusing on the most significant Q&A exchanges.Q: Mark, you hired the senior leadership of your AI labs about a year ago. Could you just give us your thoughts on how the lab's performing and do you think the street will really see an uptick in product velocity?
A: (Mark Zuckerberg, CEO) I'm quite happy with the trajectory. We've released impressive models on our early scaling ladder and are scaling much larger ones. The sustainable advantage comes from the data flywheel—learning from user behavior to improve products. We are likely the best company in the world at scaling experiences to billions of people, which is a durable advantage for our personal and business agent work.
Q: Mark, you've talked about the big pipeline for new products across consumer and business agents, API tools, and compute rental. Which of these do you expect to scale first in '26 and '27 to showcase quantifiable material ROIC for investors?
A: (Mark Zuckerberg, CEO) A substantial amount of compute goes to training models, but the rest spans improving our core business, new consumer products, the API, business agents, developer tools, and selling compute directly. We have many offers for compute at a meaningful premium. While selling intelligence will have higher margins, we expect meaningful growth in all these areas and will have more to share soon.
Q: When you frame up the enterprise opportunity, how much of that opportunity do you think is available to you today based on what you've built out in terms of go-to-market strategy as extensions of the advertising business?
A: (Mark Zuckerberg, CEO) The enterprise opportunity is a combination of extending our current business (selling to marketers via business agents on messaging apps) and building new muscles. The former is a natural extension of our partnerships with millions of advertisers. The latter includes coding and productivity tools for larger businesses, which is a somewhat different muscle but a very large opportunity. The enterprise opportunity is the sum of compute, API, productivity, and business agent services.
Q: Mark, everyone's got a story of someone using AI to write code, but then they tell their parents they're using it wrong. How do you think about whether consumer adoption can close this AI utility gap?
A: (Mark Zuckerberg, CEO) Some things have already broken through, like coding agents. Our bet is that consumer personal agents will be an extremely important market. For billions of people, it needs to just work out-of-the-box, unlike the fiddling required for developer tools. This plays to Meta's strengths in building consumer products for billions and scaling infrastructure. We are very excited about this and will ship something soon.
Q: Susan, you talked about how LLMs are increasingly capable of delivering ranking and recommendation gains. Can you talk more about the roadmap here and how far along you are?
A: (Susan Li, CFO) We see further headroom to improve recommendations into 2027. We will make recommendations more personalized by advancing our models to capture user interest more precisely. We are improving data infrastructure to train on more data and scaling up user interaction sequences. We are also incorporating LLMs for deeper content understanding and using LLM-based agentic approaches to transform the recommendation system, which also makes our engineers more productive.
Q: Mark, you've received many offers to monetize your compute externally, but you are also purchasing capacity from third parties. Can you help us understand the differences?
A: (Mark Zuckerberg, CEO) The high-level observation is there is nowhere near enough compute for all demand. We have a large number of offers for our compute, but also many valuable internal uses. The trade-off is monetizing today vs. developing future assets. It would be foolish to sell all compute for a short-term profit when we can build intelligence on top of it, which compounds the value. We are using capital to build compute, confident we can monetize it directly or through intelligence across our core business and new products.
Q: Mark, sticking with the AI lab. Muse Spark 1.1 is close to the Pareto frontier but at the lower cost end. Could you talk about competing at both ends and the role of open source?
A: (Mark Zuckerberg, CEO) We are climbing the scaling ladder. We want more advanced models for hard problems and efficient models for the vast majority of prompts. Both are important. On open source, we've always felt it's important for the ecosystem and creates positive feedback loops. We plan to do a mix of open and closed models. We wanted the MSL team to be uninhibited in building the most intelligent models, and we expect to get back to releasing some open source models soon.
Q: Mark, just to touch on the last point again around open weight models. Does that change Meta's view of developing closed proprietary frontier models?
A: (Mark Zuckerberg, CEO) No. Open source models are not as strong as frontier models. Meta is a full-stack technology company. Having sovereignty over building our own models is critical for the stack. For discerning customers, open source is important for control and trust, but building our own models is a critical part of our advantage. Building full-stack models specific to our use cases (like Instagram recommendations or personal agents) is where the durable advantage lies.
Q: Susan, you noted that you plan to maximize '26 and '27 capacity. Is that a demand or supply comment?
A: (Susan Li, CFO) It's both. We are demand-constrained today and expect to be for the foreseeable future, with many ROI-positive places to put compute. Second, there is uncertainty over long-term constraints on building capacity. Beyond '27, the world will evolve a lot. For '28, we are focusing on flexibility—having land and power—but making actual decisions on chips and big-ticket items further in the future.
For the complete transcript of the earnings call, please refer to the full earnings call transcript.
This stock alert was generated using automated technology and GuruFocus financial data to provide readers with timely and accurate market reporting. This content was reviewed by GuruFocus editorial team prior to publication. Please send any questions or comments about this story to [email protected].
