Key Points
Financial Highlights (Q2 FY2026, reported August 6)
- Revenue: $696.1M, up 36% YoY
- Gross margin: 73.1% — first sequential improvement in 8 quarters
- Non-GAAP operating income: $96.1M, up 33% YoY
- Free cash flow: $56.4M, up 69% YoY
- RPO (Remaining Performance Obligations): $2.732B, up 38% YoY
- Large customers (>$100K annual spend): 4,698, up 27% YoY; net adds of 986 over the trailing 12 months, with record net additions at every tier ($100K, $500K, $1M, $5M+)
- Dollar-based net retention: 120%, +2pp QoQ, +6pp YoY
- New paying customers: 80,000+, up 74% YoY
- Net loss: $205.7M, more than 2x YoY (driven by agent-era CapEx)
- Full-year revenue guidance raised: $2.864B–$2.870B (~32% YoY growth)
- Geographic mix: US 51% (+41%), EMEA 27% (+30%), APAC 14% (+32%)
- Nov 2025: Team projected non-human traffic would exceed human traffic in H2 2027
- Mar 2026: Revised estimate moved up to H1 2027
- May 2026: Data confirmed non-human traffic had already surpassed human traffic
- Cloudflare.pay — a payment layer purpose-built for agent traffic, charging fractions of a cent per request routed through Cloudflare's network
- ADLC (Agent Development Life Cycle)
- Agents platform
- Cloudflare official Q2 2026 earnings call transcript (Yahoo Finance / Quartr): https://finance.yahoo.com/quote/NET/earnings/NET-Q2-2026-earnings_call-662709.html
- Zhitong Finance Chinese-language earnings analysis: https://www.zhitongcaijing.com/content/detail/1477439.html
- IT Home August 8 coverage (revenue, loss, Prince quotes): https://www.ithome.com/0/987/438.htm
- Biggo Finance earnings data summary: https://finance.biggo.com/news/US_NET_2026-08-06
Humans Will Be a Rounding Error
On the Q&A call, CEO Matthew Prince disclosed a striking trajectory for non-human traffic:
Prince's quote (translated): *"In other words, humans will be a rounding error on the Internet, not because human traffic goes down, but that's just how fast we're seeing non-human traffic grow."*
Aggressive forecast: If trends continue, non-human traffic will be 1,000x human traffic within five years.
The driver is behavioral: a human shopping for a camera checks ~5 retailers; a single AI agent may query 5,000 sites. With millions of users running agents simultaneously for price comparison, research, and summarization, traffic scales exponentially. Cloudflare added ~2 million net developers in Q2 alone — more than all of 2025 combined.
Cloudflare's Repositioning as the Agent Internet Backbone
Three pieces of evidence show Cloudflare is repositioning itself as foundational infrastructure for the agent era:
1. Agents Week — full stack launch. Cloudflare simultaneously shipped:
Prince framed the shift directly: *"The business model of the Internet for the last 27 years has been largely defined by advertising and really defined by Google. The business model of the next 27 years of the Internet is going to be very different."* The implication: replacing the ad-defined model with per-request billing.
2. 80%+ of major AI companies are Cloudflare customers. Repeated by Prince on the Q&A, this signals that every major LLM provider and agent platform's outbound traffic flows through Cloudflare — positioning it as both the gateway and billing layer for the agent era, analogous to Google's role in search advertising circa 2003.
3. CapEx at 14–15% of revenue (~$430M) — low for a cloud provider. Prince's rebuttal to "why not build GPU data centers at hyperscaler scale?":
> *"We know what the typical kind of inference loads are at the hyperscalers, and their GPU utilization. It's super low. If we can, and in some cases, get 10 times as much utilization out of every CapEx dollar, then we can play these games and continue to grow revenue."*
Cloudflare's moat is in scheduling, caching, and protocol handling — not raw GPU capacity. Its model distributes workloads across existing infrastructure for 10x efficiency per dollar.
Sub-Arguments
AI traffic is not neutral. Prince split AI traffic in two: "For some of it, we actively block. We don't charge customers extra because we think it's part of being a security company." Malicious scraping of ad-dependent media is blocked by default. This clarifies Cloudflare's dual role: agent gateway *and* copyright moat.
Losses are widening — but strategically. The $205.7M net loss reflects infrastructure investment for the agent era, not operational weakness. Free cash flow grew 69% YoY, and full-year revenue guidance was raised.
Geographic mix is balanced. US 51%, EMEA 27%, APAC 14% — no single-region dependency.
Strategic Implications
1. The agent internet's business model is being forcibly defined. The shift from advertising to per-request billing carries execution risk: agent traffic (5,000 sites vs. 5) is far more complex than human traffic. Cloudflare must prove it can serve this volume without quality loss. The 10x CapEx utilization story is Prince's answer to "why not build GPU data centers?" — the moat is in scheduling, caching, and protocol handling, not bare metal.
2. "Humans are a rounding error" is a roadmap, not rhetoric. A 1,000:1 ratio in five years implies continued product tilt toward agents: CAPTCHA replaced by agent credentials; Bot Management evolving from "block bad bots" to "manage agents"; CDN routing shifting from "page acceleration" to "tool-call acceleration." For the first time, Cloudflare's pivot narrative has become numbers.
3. Practical implications for agent builders. If your product is an agent, the cost of calling third-party sites/APIs will become a core operational metric within three years. If Cloudflare.pay succeeds, domestic cloud providers (Alibaba Cloud, Tencent Cloud edge products) will likely replicate the model. Optimizing per-request cost may matter more than optimizing per-request latency.