Cloudflare Bets on AI Agent Internet, Maintains Neutral Rating

Overview

This research report is Bernstein's commentary on Cloudflare's 2026 Investor Day. The report points out that Cloudflare is fully shifting its strategic focus to 'AI First', committed to becoming the default cloud platform for the emerging 'Agentic Internet'. Although management presented a clear technical vision and updated long-term financial goals, analysts believe the current stock price fully reflects expectations, and the company faces fierce market competition and execution risks, thus maintaining the Market-Perform rating and $136 target price unchanged.

Core views

AI Agent Strategy and Architectural Advantages: Cloudflare divides its development path into four stages (Act I-IV), currently in the fourth stage centered on content monetization and AI agents. The company launched agent-centric Cloudflare OS, including AI Agent Workspace, centrally managed context layer, security record system, and LLM model optionality. The report believes that compared to the microservice-based architecture of traditional hyperscale cloud vendors, Cloudflare's edge-native architecture has structural performance and cost advantages when handling highly dynamic, personalized AI agent workloads. About 80% of leading AI companies are using Cloudflare, validating its relevance in the AI ecosystem. Security Challenges and Developer Platform: Security is seen as the most urgent challenge and opportunity in the near term. AI compresses the time from vulnerability discovery to exploitation to seconds; enterprises will focus on vulnerability identification and mitigation in the next two years. Cloudflare positions itself as a pre-deployment and real-time security layer through partnerships with model providers like Anthropic. In terms of developer platform, AI-driven code generation is surging, but due to context window limits and tool call efficiency issues, AI agents still need to write and execute code themselves, which reinforces the demand for lightweight, low-cost computing primitives, benefiting Cloudflare's Serverless and Greenfield deployments. Competitive Concerns and Market Transformation: Although the strategy is clear, the report highlights significant competitive risks. Especially the Vercel+Next.js combination constitutes a strong threat. In addition, edge network/agent-based methods may have disadvantages; the tight coupling of platform and network raises customer concerns about vendor lock-in, and mainstream code suggestion tools often do not prioritize Cloudflare when designing architecture. In terms of commercialization, the company is transitioning from broad sales to precise enterprise-level sales; partner revenue share has increased from about 20% to about 30%, and AI is used to optimize internal sales processes. Financial Goals and Unit Economics: Management updated the long-term financial framework, targeting $5 billion ARR by the end of 2028, reaching Rule of 50 (growth rate + margin ≥ 50%) in 2027, and achieving GAAP profitability in 2028. Current unit economic margin is about 44%. Although free tier investment and new product mix put pressure on gross margin, management emphasizes this is in exchange for long-term scale, similar to the early evolution path of Microsoft's cloud business. Notably, thanks to productivity improvements brought by AI, the company raised its long-term operating margin expectation by about 10 percentage points.

Analysis framework

Analysts adopted an analysis framework of 'Strategic Narrative Verification + Competitive Reality Check'. First, they sorted out the technical evolution logic of the 'Agent Internet' proposed by the company at the Investor Day, evaluating the theoretical advantages of its architecture relative to traditional cloud vendors; second, they introduced an external competitive perspective (such as Vercel, code tool ecosystem) to hedge against management's optimistic narrative, identifying potential execution friction points; finally, combined with the updated unit economic model and long-term financial guidance, they judged whether the current valuation has overdrafted growth expectations. This analysis method respects the company's long-term vision while maintaining an independent risk pricing perspective.

Methodology notes

• Valuation MethodDCF Discounted Cash Flow

  Blended Valuation

  The report uses the average of DCF (WACC 10%, perpetual growth rate 3%) and forward P/S multiple (16x Next 5-8Q Sales) as the target price. This method combines absolute value assessment and relative market pricing, often used for high-growth but not yet fully profitable SaaS companies, to balance intrinsic value and market sentiment fluctuations.

• Industry/Industrial Analysis FrameworkSubstitution Effect Analysis

  Architectural Paradigm Shift

  The report analyzes how AI agent workloads migrate from traditional centralized microservice architectures to edge-native architectures. When the requirements of new technology paradigms (such as Agent) for latency, personalization, and cost exceed the optimization scope of old architectures, a substitution effect occurs, which explains Cloudflare's differentiated opportunities relative to traditional hyperscale cloud vendors.

• Company Fundamentals and Financial FrameworkFree cash flow analysis

  Unit Economics and Long-term Margin Modeling

  Focuses on unit economic margin (~44%) rather than just looking at overall gross margin, because SaaS companies often sacrifice short-term gross profit due to strategic investments (such as free tiers, new products) during expansion. By analyzing unit-level profitability and operational leverage brought by AI (SG&A/R&D efficiency improvements), long-term margins and free cash flow generation capabilities can be predicted more accurately.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Cloudflare Inc (NET.US)

  Core coverage target, direct carrier of AI agent strategy

  Strengths

  Edge-native architecture suitable for AI agent workloads; ~80% of leading AI companies are already users; deep integration of security and network platform; healthy unit economics (~44% margin)

  Weaknesses

  Disadvantaged in developer tool chain compared to competitors like Vercel; strong coupling of platform and network raises lock-in concerns; gross margin under short-term pressure; current valuation is high

  Comparison

  Compared to traditional hyperscale cloud vendors, has more cost and performance advantages on dynamic AI workloads; compared to Vercel, owns more complete security and network infrastructure, but lags in front-end development experience and code tool ecosystem integration

  Risks

  Cloud growth slowdown; sales model transformation execution risk; low visibility of consumption-based revenue; Act II/III product demand validation failure; hyperscale cloud vendors entering core CDN and security fields

Key data

• Target Price136.00 USDMaintained unchanged, based on average of DCF and 16x P/S
• Current Stock Price236.13 USDAs of close on June 9, 2026
• 2028 ARR Target$5 billionLong-term revenue milestone set by management
• Unit Economic Margin~44%Current unit-level profitability indicator
• Partner Revenue Share~30%Significantly increased from previous ~20%, effectiveness of channel transformation
• 2026E Adjusted EPS1.33 USDBernstein forecast value

Impact & implications

For Cloudflare, the AI agent strategy is a key bet for its leap from a CDN and security vendor to a full-stack cloud platform. If successful, its edge architecture will become the standard runtime environment for AI applications, opening up a TAM (Total Addressable Market) far exceeding traditional cybersecurity. However, gross margins will remain under pressure in the short term, and it must prove its developer experience and commercial monetization capabilities under the夹击 of emerging competitors like Vercel. For investors, this means the company is in a validation period where high potential coexists with high risks; current valuation includes high expectations, requiring close attention to the substantive conversion of AI-related revenue and continuous improvement of unit economics.

Risks

• Macroeconomic downturn affects cloud spending
• Sales model transformation execution falls short of expectations
• Predictability of consumption-based revenue is low
• Act II/III new product demand validation failure
• Hyperscale cloud vendors encroach on core CDN and security business
• Current valuation level is high, leaving little room for error
• Fierce competition from emerging platforms like Vercel+Next.js
• Edge agent architecture may lead to some customer churn due to avoidance of vendor lock-in

What to watch

• Actual adoption rate and revenue contribution of AI agent-related products
• Penetration rate of large enterprise customers and progress of platform integration orders
• Whether unit economic margin can expand with AI productivity improvements
• Market share changes of developer platform in Serverless and Greenfield projects
• Realization of incremental demand for security business under new AI attack surfaces
• Progressive achievement of 2027 Rule of 50 target