Akamai delivered a strong Q3 2025, beating expectations with revenue of $1.055 billion (up 5% YoY) and Non-GAAP EPS of $1.86 (up 17% YoY). Growth was driven by the acceleration of Cloud Infrastructure Services (CIS), which grew 39% YoY to $81 million, and robust Security revenue of $568 million (up 10% YoY). Management highlighted the strategic launch of 'Akamai Inference Cloud' in partnership with NVIDIA to address the burgeoning demand for distributed AI inference at the edge. While Delivery revenue continued to decline (-4% YoY), pricing pressures are moderating and the segment is stabilizing. Looking ahead, the company raised full-year EPS guidance to a range of $6.93 to $7.13, reflecting confidence in operating leverage and the high-growth potential of its security and compute portfolios.
| Metric | Value | Change |
|---|---|---|
| Revenue | $1.055 billion | +5% YoY |
| Non-GAAP EPS | $1.86 | +17% YoY |
| Cloud Infrastructure Services (CIS) Revenue | $81 million | +39% YoY |
| Security Revenue | $568 million | +10% YoY |
| Delivery Revenue | $306 million | -4% YoY |
| Non-GAAP Operating Margin | 31% | Improved YoY |
| CapEx | $224 million | 21% of Revenue |
Akamai is making a definitive strategic bet on AI Inference at the edge with the launch of 'Akamai Inference Cloud' in partnership with NVIDIA. By leveraging NVIDIA's Blackwell GPUs across Akamai's distributed network of 4,000+ points of presence, Akamai aims to solve the latency and scalability issues of centralized AI inference. This moves Akamai beyond its CDN roots into a critical layer of the AI stack, targeting use cases like personalized media, real-time gaming, and robotic control that require response times in the tens of milliseconds.
Cloud Infrastructure Services (CIS) is rapidly becoming a primary growth engine, accelerating to 39% YoY growth in Q3. A major validation of this strategy is that all three top U.S. cloud providers (hyperscalers) are now utilizing Akamai's CIS for distributed compute needs, such as ad logic and media workflow. This 'coopetition' dynamic underscores the unique value of Akamai's edge platform, which hyperscalers cannot easily replicate due to the physical logistics of edge distribution.
The Security business is demonstrating resilience and high-growth potential, specifically within API Security and Zero Trust segmentation. High-growth security products grew 35% YoY, with API security specifically expected to exit 2025 with a $100 million run rate. Management is effectively cross-selling these solutions, evidenced by a $7 million API security deal with a major European bank and a $42 million expansion with a global software company.
Akamai is actively transforming its go-to-market strategy by hiring 'hunters' and specialists to drive new business in Compute and Security. This structural change is designed to support larger, multi-year deals and improve attach rates across the portfolio. Management indicated that this transformation will be largely complete by early 2026, setting the stage for improved sales productivity and revenue acceleration in the coming year.
Capital allocation remains disciplined, balancing heavy investment in CIS CapEx with shareholder returns. The company executed the largest buyback in its history in the first half of the year ($800 million) but paused in Q3 to preserve flexibility. This opportunistic approach suggests management is prioritizing strategic investments in AI infrastructure while remaining committed to returning cash when valuations are attractive.
The Delivery business, while stabilizing, continues to contract, with revenue down 4% YoY in Q3. Management attributes this to legacy headwinds and pricing pressure, noting that 'Other Cloud Applications' (OCA) within Compute remains a drag. While pricing declines have moderated, the secular shift away from traditional CDN toward higher-value compute services means Delivery may remain a drag on total growth for the foreseeable future.
Capital Expenditures surged to $224 million (21% of revenue) in Q3, driven by the build-out of the Inference Cloud and CIS infrastructure. While management asserts a 1:1 revenue-to-CapEx ratio over time, the upfront investment creates near-term cash flow pressure and execution risk. If demand for Inference Cloud does not materialize as quickly as deployed capacity, gross margins could face temporary headwinds.
Guidance for Q4 implies a potential deceleration in EPS growth, with a wide range of $1.65 to $1.85 compared to the $1.86 achieved in Q3. Management cited seasonal sales commissions and the unpredictability of holiday traffic as factors. The wide range introduces uncertainty, particularly regarding the performance of the Delivery segment during the critical holiday season.
The 'Other Cloud Applications' (OCA) segment within Compute remains a headwind, masking the true growth of the core CIS business. Management noted that OCA is 'pretty small, pretty stable' but acts as a drag on the total Compute growth rate, keeping it around 15% despite CIS growing at nearly 40%. Investors must look past the top-line Compute number to see the underlying strength of CIS.
Overall: Management exhibited a high degree of confidence and enthusiasm, particularly regarding the company's strategic pivot towards distributed cloud computing and AI. Dr. Leighton was visionary and assertive about Akamai's unique positioning at the edge, while CFO Ed McGowan remained grounded and precise regarding financial execution and capital allocation. The tone shifted from defensive in past years about legacy delivery to offensive about new growth vectors like AI inference.
Confidence: HIGH - Management used definitive language regarding the 'acceleration' of CIS and 'strong demand' for security. The specific disclosure of major contract wins (e.g., hyperscalers, global banks) and the bold launch of the Inference Cloud signal strong conviction in their strategy.
$1.065 billion - $1.085 billion (+4% to +6% YoY)
$1.65 - $1.85
+4% to +5% constant currency
$6.93 - $7.13
29% - 30%
Hedging & Uncertainty: Management generally used confident, forward-looking language ('We believe,' 'We are positioned') but employed specific hedges regarding the timing of new revenue streams. Phrases like 'first inning' and 'very good chance we could accelerate' indicate high potential but acknowledge early stages. Ed McGowan used more temporal hedging regarding guidance, stating 'we don't give guidance by quarter' and 'it's hard to say what the retail season will look like,' which serves to manage analyst expectations for Q4 volatility.
"Inference has become the most compute-intensive phase of AI, demanding real-time reasoning at planetary scale." - Jensen Huang (quoted by F. Leighton, CEO)
"We're in the first inning of a very exciting game." - F. Leighton, CEO
"There's a very good chance we could accelerate growth in our CIS business next year." - Ed McGowan, CFO
"We see the same need for performance, scale and security playing out again with AI inference today." - F. Leighton, CEO
"Our goal at Akamai is to have five 9s of reliability." - F. Leighton, CEO
Analyst Sentiment: Analysts were highly inquisitive about the monetization timeline of the new AI Inference Cloud and the sustainability of CIS growth. Questions focused heavily on the competitive dynamics with hyperscalers and the margin profile of the new GPU infrastructure.
Management Responses: Management was deft in deflecting concerns about hyperscaler competition by emphasizing 'coopetition' and the unique physical distribution of Akamai's network. They were transparent about the 'timing' differences between CapEx deployment and revenue recognition for AI, while maintaining a bullish long-term outlook.
AI Inference Cloud & NVIDIA Partnership: Analysts sought details on the commercial readiness and customer adoption of the new Inference Cloud. Management emphasized it is in the 'first inning' with strong initial interest from sectors like media and sports broadcasting.
CIS Growth & Hyperscaler Adoption: Multiple questions focused on the 39% growth in CIS and the significance of winning all three major cloud providers as customers. Management clarified that this is due to performance needs (latency) rather than hyperscaler capacity constraints.
Delivery Business Stability: Analysts asked if the stabilization in Delivery revenue is sustainable. Management attributed the improvement to moderating price declines and fewer competitors in the market, though they remained cautious on Q4 seasonality.
Security Attach Rates: There was interest in how AI inference drives security sales. Management highlighted the need for 'AI firewalls' and API security to protect these new inference models, seeing a natural attach rate there.
Akamai is successfully executing a pivot from a legacy CDN to a high-growth distributed cloud security and compute provider. The Q3 results demonstrate that the core business is stabilizing while the new growth engines (CIS and Security) are accelerating. The launch of the Akamai Inference Cloud with NVIDIA represents a significant differentiator, positioning Akamai as a critical infrastructure provider for the next phase of AI (Inference at the Edge). The company is balancing heavy investment in this future growth with strong shareholder returns (buybacks) and expanding operating margins. While Delivery remains a headwind and CapEx is elevated, the risk/reward is favorable given the 39% growth in CIS and the $100M run-rate emerging in API security. Akamai is uniquely positioned to win in the 'Edge AI' market.
Management highlighted a structural shift in AI from training to inference, noting that 'Edge Inference leads the market' and is projected to grow at the highest CAGR. This suggests a long-term capex cycle towards distributed compute.
FX had a positive $8 million impact YoY in Q3 and is expected to positively impact full-year revenue by $13 million, providing a slight tailwind to reported figures.
Management noted 'fewer competitors in the marketplace' for Delivery services, as some have exited or gone out of business. This is contributing to the stabilization of pricing in the legacy segment.