DalalBytes verdict
Akamai 78, Cloudflare 76.

Both companies reported on November 5, 2026, and the two prints pulled in opposite directions. Cloudflare delivered its best quarter ever: revenue up 39%, growth accelerating at a $3B run rate, guidance raised for the third time this year. Akamai delivered a modest quarter and an immodest contract book: 6% growth today, and a signed $14B of cloud commitments with published returns and a dated 2027 ramp. The scorecard rewards what is delivered, so Cloudflare gains three points of ground. It does not reward hope, so Akamai keeps the lead on cash reality and price.

The valuation gap remains the story. Cloudflare trades near 35x trailing sales; Akamai trades near 2.7x guided 2026 sales and below 1.7x its guided 2028 revenue. Our decomposition, set out in full in the valuation section, finds that growth and margin quality justify a large premium for Cloudflare, plausibly a 4x to 6x multiple ratio. They do not justify the roughly 11x ratio that survives into 2028, when both companies are guided to grow near 30%. The residual prices Akamai's signed book at zero and Cloudflare's next decade as certain.

What would change Akamai's score. CIS revenue tracking at or above the $1.1B low end of the 2027 guide by mid-2027 raises it; a second delivery delay, or evidence against the published 13% IRR frame, cuts it.

What would change Cloudflare's score. Growth holding above 35% while stock compensation falls toward 15% of revenue raises it; deceleration below 30% growth at 35x sales cuts it.

Frequently asked questions

Who wins between Akamai and Cloudflare?

Akamai, narrowly: 78 to 76 on the DalalBytes ten-dimension scorecard, rebuilt after both Q3 2026 prints. Cloudflare takes growth, portfolio, and moat. Akamai takes AI contract scale, cash discipline, and valuation. The October scores were 79 and 74; Cloudflare's accelerating quarter closed most of the gap.

How much of the valuation gap between Cloudflare and Akamai is justified?

About half. Growth justifies most of today's 13x sales-multiple gap because Cloudflare grows 39% and Akamai grows 6%. But on 2028 guided numbers, where both grow near 30%, an 11x gap remains. Margin quality and platform depth defend perhaps 4x to 6x of it. The rest prices Akamai's signed book at zero.

What is the biggest risk for each company?

For Cloudflare, the price: 35x trailing sales and stock compensation near 21% of revenue leave no room for a growth stumble, and gross margin has slipped three points in a year. For Akamai, execution: a $2.4B to $2.6B capex quarter, one customer at $11.6B of a $14B book, and a ramp that must land on schedule.

What did the two Q3 2026 prints actually say?

Q3 2026AkamaiCloudflare
Revenue$1.122B, up 6% (up 7% constant currency)$782.5M, up 39%
Earnings per share$1.77 non-GAAP; $0.77 GAAP, down 21%$0.38 non-GAAP; -$0.10 GAAP
Operating margin25% non-GAAP; GAAP operating income down 35%17.5% non-GAAP, a record; GAAP operating loss of $67.9M
Gross marginAbout 56% GAAP on the last full year71.2% GAAP, down from 74.0% a year earlier
Operating cash flow$433M, 39% of revenue$215.4M, 28% of revenue
Capex in the quarter$464M, 41% of revenueAbout $133M, derived from cash flow less free cash flow
Free cash flowAbout -$31M, derived$82.1M, 10% of revenue, down from 13%
Guidance moveFY2026 trimmed at the edges; 2027 and 2028 guided for the first timeFY2026 raised for the third time this year
  • Akamai's quarter was in line and its guide was the news: 2027 total revenue growth of 17% to 22% in constant currency, 2027 CIS revenue of $1.1B to $1.3B, and 2028 revenue above $7.0B.
  • Cloudflare's quarter was the news: 39% growth was its fastest in nearly four years, at a revenue run rate above $3B annualized.
  • Both stocks had fallen hard into the prints in a broad AI selloff. Akamai closed the print day down 9.4% before its release, then rose 6.7% on November 6. Cloudflare rose about 10% on November 6 after its release.

Tale of the tape

Akamai (AKAM)Cloudflare (NET)
Price, November 6, 2026 close$83.09$268.75
Market cap$11.94BAbout $96.3B on 358.2M shares
Revenue, last twelve monthsAbout $4.4B$2.73B, settled in the caveats below
FY2025 revenue$4.21B, up 5%$2.17B, up 30%
FY2026 guided revenue$4.44B to $4.465B$2.963B to $2.964B
Price to FY2026 guided sales2.7x32.5x
Price to trailing salesAbout 2.7x35.2x
GAAP profitable, trailing yearYesNo; GAAP EPS about -$0.29
Free cash flow$699M in FY2025; negative through the 2026 buildAbout $322M trailing, derived from releases
Cash and securities$4.2B at September 30, 2026$5.13B at September 30, 2026
Debt$7.56B of converts at June 30, 2026$4.44B of converts at September 30, 2026
Network4,400+ locations, 130+ countries330+ cities, 120+ countries
Share count trendDown 18.5% from 2015 to 2025; buyback paused in 2026Rising about 2% a year on stock compensation
CEOTom Leighton, co-founderMatthew Prince, co-founder
Founded19982009

Where does each company's growth come from?

Akamai reports three segments. Cloudflare reports one segment and discloses no product revenue, so its mix has to be read through customer and bookings metrics. The two tables below use each company's own disclosure standard, applied symmetrically.

Akamai segmentQ2 2026Q3 2026Q3 year over yearShare of Q3 revenue
Security$604M$622MUp 9%, up 10% constant currency55%
Delivery and other cloud applications$396M$391MDown 3%35%
Cloud Infrastructure Services$99M$109MUp 34%, up 35% constant currency10%
  • Security is the funder: 55% of revenue, growing at the low edge of double digits, with API Security and Guardicore segmentation leading.
  • Delivery is stabilizing rather than collapsing: down 3% in Q3 against down 6% in Q2, with pricing described on the call as holding better than it has in a long time.
  • CIS is the swing factor: 10% of revenue today, guided to $1.1B to $1.3B in 2027, which would make it roughly a quarter of the company.
  • International revenue was $563M in Q3, half the total, up 9% in constant currency.
Cloudflare mix metricLatest readingSignal
Large customers above $100K ARRMore than 5,000, up about 25% year over yearEnterprise mix keeps rising; 73% of revenue at Q2
Dollar-based net retention122% in Q3 2026, up from 119% a year earlierExisting customers are expanding faster
Remaining performance obligations$2.88B, up 34%; current portion up 37%Booked business supports the guide
Billings$846.2M in Q3, up about 35%Cash collections track the revenue growth

1. Which business is more diversified? Akamai 7, Cloudflare 8

  • Akamai. Three disclosed segments with different cycles: security grows, delivery shrinks slowly, cloud infrastructure ramps from contract.
  • Akamai. Half of revenue is international, and no single customer was 10% of revenue before the Anthropic signing.
  • Akamai. The weakness is forward concentration, not current mix: one customer is $11.6B of the $14B signed CIS book.
  • Cloudflare. One segment and no product revenue disclosure, a standing transparency minus.
  • Cloudflare. Breadth shows in the customer base instead: more than 5,000 large customers, millions of free users, and 122% net retention spreading growth across the base.
  • Cloudflare. When every cohort grows near 30% or better, diversification matters less than durability of the funnel.

2. Which product portfolio is stronger? Akamai 7, Cloudflare 9

  • Akamai. Enterprise-grade depth: App and API Protector, Bot Manager, Account Protector, Guardicore Segmentation, Prolexic, Enterprise Application Access, Edge DNS.
  • Akamai. The compute stack is real and widening: CIS, Inference Cloud, managed Kubernetes, video processing, built on the Linode base.
  • Akamai. It wins RFPs. It still does not win developers' hearts, and there is no self-serve funnel of consequence.
  • Cloudflare. The broadest developer platform at the edge: Workers, Pages, R2 with zero egress fees, D1, KV, Queues, Vectorize, Durable Objects.
  • Cloudflare. Cloudflare One assembles a full SASE suite, and Pay Per Crawl, the HTTP 402 mechanism from June 2026, lets publishers charge AI crawlers at a rate of about a billion payment-required responses a day.
  • Cloudflare. A generous free tier makes the top of the funnel enormous, and each product cross-sells the next.

3. Who is winning edge compute and AI? Akamai 9, Cloudflare 7

  • Akamai. More than $14B of multi-year CIS commitments are signed, against no disclosed Cloudflare contract of comparable scale.
  • Akamai. The Anthropic commitment is $11.6B over seven years, with a potential $9B expansion, dedicated CPU capacity, and managed support.
  • Akamai. Management published the book's economics on the Q3 call: about $2.2B of annual revenue at full ramp, about 75% adjusted EBITDA margin, about 30% non-GAAP operating margin, about 13% after-tax IRR, and about 15% average annual adjusted ROIC. These are management figures, graded at delivery.
  • Akamai. Two of the first three large deals are already billing; the third is in final testing. Space and power for the Anthropic build are secured, and about 70% of the hardware is in hand.
  • Akamai. The AI Grid orchestration layer, live since March 2026 as the first global implementation of NVIDIA's reference design, routes each request across edge, regional, and core tiers by cost, latency, and throughput.
  • Akamai. Semantic caching, matching requests by meaning rather than exact wording, is disclosed as an orchestrator capability. No hit-rate metrics have been published, and we score it as a capability, not a result.
  • Akamai. Inference Cloud launched in October 2025 across about 20 initial GPU regions; well over 90% of the servers now being deployed are CPUs with useful lives of seven to ten years.
  • Cloudflare. Workers AI runs serverless inference on GPUs in 180+ cities across 89+ open models, priced per request, the most elegant developer inference fabric in the industry.
  • Cloudflare. Q3 brought an $11M pool-of-funds Workers contract from a rapidly growing AI company, evidence that agentic workloads are starting to sign real money.
  • Cloudflare. AI Gateway sits in front of third-party models with exact-match caching served from the edge, dynamic routing, and unified billing with Workers AI since August 2026.
  • Cloudflare. Semantic caching at the gateway has been described as planned since 2024 and remained unshipped in late 2026 documentation reviews; the shipped cache is exact-match only.
  • Cloudflare. Management said more than half of the traffic on its network was already nonhuman in Q2 2026, which positions the network where agent traffic actually flows.

The honest summary: Akamai sells contracted enterprise capacity and has the signatures to prove it. Cloudflare sells developer-adopted optionality and has the traffic to prove it. Only one of them has signed AI revenue at scale, and only one of them owns the layer where the next decade of agent traffic will be routed. The scores reflect signed scale today.

4. Whose network is bigger? Akamai 9, Cloudflare 8

  • Akamai. 4,400+ locations in 130+ countries and 700+ cities, with 345,000+ edge servers and over a petabit per second of throughput.
  • Akamai. Twenty-five years of embedding inside ISP networks is not replicable on a budget or a timeline.
  • Cloudflare. 330+ cities in 120+ countries, 13,000+ interconnections, roughly a fifth of all websites behind the network, and billions of threats blocked daily.
  • Cloudflare. Every location runs the full stack, so capacity is uniform and instantly available to any product.
  • The nuance. Akamai's count includes thousands of small embedded clusters, often a rack or two inside someone else's facility, that cannot run heavy compute. Cloudflare's fewer locations are each a real data center presence. Raw reach goes to Akamai; uniformity goes to Cloudflare.

5. Who grows faster, and for how long? Akamai 6, Cloudflare 10

Revenue growthAkamaiCloudflare
FY2023Not restated hereUp 33%
FY2024Not restated hereUp 29%
FY2025Up 5%Up 30%
Q3 2026Up 6%Up 39%
FY2026 guideAbout up 6%Up about 37%
2027Up 17% to 22% guided, constant currencyEstimates near $3.7B, about up 25% on the raised guide
2028Above $7.0B guided, about up 30% impliedAbove $5B targeted by management
  • Cloudflare. Growth is accelerating at scale: 36% in Q2, 39% in Q3, the fastest in nearly four years, with the full year guide raised three times in 2026.
  • Cloudflare. Bookings confirm it: RPO up 34%, current RPO up 37%, net retention up to 122%.
  • Akamai. Delivered growth is still 6%, and the score does not pay for guidance in full.
  • Akamai. The acceleration is contracted and dated rather than hoped for: the 2027 range is built on signed deals only, and management says the low end already assumes further delays.
  • Akamai. That earns one point over October's 5, not four. The remaining points are earned at delivery, the same standard applied to Cloudflare's guidance.

6. Who makes money, and who makes cash? Akamai 7, Cloudflare 6

MarginAkamaiCloudflare
GAAP gross marginAbout 56% in FY202571.2% in Q3 2026, down from 74.0%
Non-GAAP operating margin25% in Q3 202617.5% in Q3 2026, a record high
GAAP operating marginPositive; about 11% in FY2025, compressing in the buildNegative 8.7% in Q3 2026
Stock compensation burdenBuybacks have exceeded it for a decade22.2% of revenue in Q3; 20.9% across nine months
Free cash flowAkamaiCloudflare
FY2025$699M, about 17% of revenue$261M, about 12% of revenue
Q3 2026About -$31M, derived$82.1M, 10% of revenue
Trailing twelve monthsNegative; FY2026 capex is guided at $3.46B to $3.66BAbout $322M, derived from the releases
Normalized viewAbout 17% of revenue before the build; the signed book targets about 75% adjusted EBITDA margin at full rampManagement targets a 30% to 35% long-term free cash flow margin; the current figure is 10% to 12%
  • Akamai. GAAP profitable throughout, with $452M of net income in 2025 and a 25% non-GAAP operating margin held in Q3 2026.
  • Akamai. The cash trough is real: Q3 free cash flow was modestly negative and Q4 capex alone is guided at $2.445B to $2.645B. We score the trough, not just the contract behind it, so this dimension falls from 8 to 7.
  • Akamai. The trough is contracted rather than speculative: the spend sits against signed revenue, and depreciation on the pre-purchased equipment begins only as revenue recognition starts across 2027 and 2028.
  • Cloudflare. Non-GAAP operating margin reached a record 17.5% in Q3, and free cash flow stayed positive while growth accelerated. This dimension rises from 5 to 6.
  • Cloudflare. The GAAP bridge remains the objection: stock compensation and related taxes were $442M across nine months, about 21% of revenue, against about $322M of trailing free cash flow. After paying employees in stock, trailing owner cash flow is negative by roughly $240M on derived figures.
  • Cloudflare. Gross margin is drifting the wrong way, 74.0% to 71.2% in a year, as Workers AI and large enterprise deals mix into revenue. Management guided to GAAP profitability in Q4 2026 on the call; it has one quarter to land.

7. Whose balance sheet and capital allocation are better? Akamai 7, Cloudflare 6

At September 30, 2026 unless notedAkamaiCloudflare
Cash and securities$4.2B$5.13B
Debt$7.56B of converts at June 30, 2026, of which $1.71B current$4.44B of converts
Net positionNet debt of about $3.0B to $3.4BNet cash of about $0.69B
Share countDown 18.5% from 2015 to 2025358.2M shares, up 1.8% in nine months
2026 capital returnsAbout $616M repurchased before the pause; $565M of authorization remainsNone; no buyback program of record
Equity overhangAnthropic warrants on 7.7M shares, about 5%, at a $111.33 strikeOngoing stock compensation issuance, about 2% a year
  • Akamai. The funding plan stated on the Q3 call is cash plus operating cash flow covering both the CIS build and the 2027 convertible maturity, with measured incremental financing as a supplement and a stated goal of holding investment grade.
  • Akamai. The 2026 buyback pause is explicitly to fund contracted demand, and it followed about $616M of repurchases earlier in the year. The pause still costs a point: leverage is up, returns are paused, and the Anthropic warrant adds a 5% overhang that vests as payments begin.
  • Cloudflare. The balance sheet is a fortress: $5.13B of cash and securities against $4.44B of converts after issuing $2.5B and repaying $1.29B during 2026.
  • Cloudflare. The leak is equity, not debt. The share count rises every year, there is no buyback to offset it, and the Q4 guidance share count of about 379M diluted shares against 358M outstanding shows the size of the dilutive overhang.

8. What does the price assume? Akamai 9, Cloudflare 4

Price to sales, November 6, 2026 closesAkamaiCloudflareRatio
Trailing revenueAbout 2.7x35.2xAbout 13x
FY2026 guided revenue2.7x32.5xAbout 12x
2027 revenueAbout 2.2x, derived from guided growthAbout 26x, on pre-print estimates near $3.7BAbout 12x
2028 revenueBelow 1.7x, guided above $7.0BBelow 19.3x, targeted above $5BAbout 11x
Price to FY2026 non-GAAP EPS12.6x on the $6.58 guide midpoint199x on the $1.35 guiden/m
Trailing · Akamai
2.7x
Trailing · Cloudflare
35.2x
2027E · Akamai
2.2x
2027E · Cloudflare
26.0x
2028 guide · Akamai
1.7x
2028 target · Cloudflare
19.3x

Price to sales on a log scale; a linear scale would erase Akamai's bars. The 2028 labels are ceilings, since both revenue figures are floors.

How much of the gap is justified?

  • Growth justifies most of the spot gap. Per point of current growth, Cloudflare costs about 0.90x sales (35.2x divided by 39% growth) and Akamai about 0.45x (2.7x divided by 6%). On today's numbers, Cloudflare is only about twice as expensive per unit of growth.
  • The gap stops narrowing when growth converges. On 2027 figures the per-point cost is about 0.9x for Cloudflare against about 0.11x for Akamai. On 2028 guided figures it is about 0.6x against about 0.06x. A roughly 10x gap per unit of growth survives the convergence.
  • Margin quality defends a real premium. Cloudflare's 71% gross margin is software-like; Akamai's roughly 56% carries the cost of owning and running a physical network.
  • Capital intensity defends more of it. Cloudflare's capex ran about 13% of revenue across the first nine months of 2026. Akamai's network model is structurally heavier even outside the current build.
  • Duration defends the rest of the defensible part. Cloudflare's funnel, 122% net retention, and expanding product surface give its growth a longer visible runway than Akamai's contracted staircase, which flattens after 2028 without new signings.
  • Our judgment. A multiple ratio of about 4x to 6x is defensible on growth duration, margin quality, and platform depth. The actual forward ratio near 11x is not. It prices Akamai's execution failure and Cloudflare's perfection at the same time.

The sum-of-the-parts cross-check on Akamai

Akamai partAnnualized revenueMultiple appliedImplied value
Security$2488M, Q3 revenue of $622M times four5.0x$12440M
Delivery and other cloud applications$1564M, Q3 revenue of $391M times four1.0x$1564M
Cloud Infrastructure Services$436M annualizedZero in this test$0
Conservative total$14004M
Market cap, November 6 close$11940M
Parts, CIS at zero
$14.0B
Market cap
$11.9B
Parts, CIS at 3x 2027 guide
$17.6B

Security at 5x sales is a deliberately plain multiple for a segment growing about 10% with the company's best margins. The market price sits about 17% below the conservative total, before CIS is worth anything at all.

  • The reverse check on Cloudflare. At 35.2x trailing sales, today's price pays in advance for roughly a decade of compounding near 30%. Management's own target, revenue above $5B by the end of 2028, still leaves the stock near 19x that year's sales.
  • The symmetric statement. Akamai's price assumes its signed future is worth little. Cloudflare's price assumes its unsigned future is worth almost everything. Both assumptions cannot be priced correctly at once, and the score gap of 9 to 4 is this page's widest for a reason.

9. Who leads better? Akamai 9, Cloudflare 9

  • Akamai. Tom Leighton co-founded the company in 1998, has run it since 2013, takes a $1 salary with nearly all pay at risk, and bought 50,000 shares in the open market in August 2025.
  • Akamai. On the Q3 call he and CFO Ed McGowan published the returns frame for the largest bet in company history and owned the $30M revenue slip in plain language. Under-promise, over-deliver, engineer.
  • Cloudflare. Matthew Prince co-founded the company in 2009 and built it from a conference launch into a roughly $96B company, and he remains the industry's clearest communicator of the agentic internet thesis.
  • Cloudflare. Q3 2026 was, by his own description and by the numbers, the company's best quarter ever, with a third guidance raise in a year.
  • Cloudflare. The debits are cultural and start at the top: a stock compensation culture that costs shareholders about a fifth of revenue, and a May 2026 restructuring of about 1,100 roles that produced $161.7M of charges across nine months.

10. Whose moat is wider? Akamai 8, Cloudflare 9

Akamai's moat is physical and contractual. The 4,400+ locations embedded inside access networks over twenty-five years cannot be rebuilt cheaply or quickly, and the $14B of signed commitments now sits on infrastructure the company already owns and operates. It is a wide moat with a static character: it deepens only as fast as the footprint converts into contracts, and the Anthropic signing is the first conversion at true scale.

The Q3 disclosure improved it in one specific way, because a customer paying for seven years of capacity on published return terms is a different asset from a customer renting month to month.

Cloudflare's moat is a flywheel. Roughly a fifth of the web sits behind the network, the free tier converts attention into distribution, the developer ecosystem makes each product cross-sell the next, and the telemetry from that traffic feeds the security products that enterprises pay for. The measurable output is the 122% net retention rate: the installed base compounds on its own. This moat does not need a signature to deepen. It deepens with usage, every day, including days when nothing is announced.

The honest contest is scarcity against compounding. Money can eventually buy locations, permits, and power, which is why Akamai's moat scores 8 rather than 9. Money cannot buy a developer default or a fifth of the web's traffic, which is why Cloudflare's scores 9. Time cuts the other way: a signed seven-year book is bankable today, while a flywheel must keep spinning at 30% growth to justify the price it currently carries. The moats are both real. Only one of them is being charged rent at 35x sales.

What is the story of these two companies?

Akamai was founded in 1998 out of MIT to solve web congestion, and it became the delivery layer of the commercial internet. The market then spent a decade repricing it as a melting asset: a fine security company growing over a shrinking CDN.

The turn came in stages. Linode, acquired in 2022, gave the footprint a real cloud. The security business became the funder. And the delivery asset the market had discounted turned out to be the physical plant that distributed AI inference needed, close to users, already permitted, already powered, already staffed by a network operations culture with twenty-five years of practice.

Cloudflare was founded in 2009 and launched with a free tier that infrastructure incumbents considered a toy. The free tier became the largest top-of-funnel in the industry: millions of properties generating telemetry, distribution, and a pipeline of future enterprise contracts. Each layer since, DNS, DDoS protection, WAF, Zero Trust, Workers, R2, has been another product sold into the same installed base through the same network. It is the cleanest land-and-expand machine in infrastructure software.

In 2026 both companies claim the same future, the agentic internet, from opposite financial shapes. Akamai arrives with contracts, capex, and a returns framework it has agreed to be graded on. Cloudflare arrives with traffic, developers, and a retention rate. The next five years grade which arrival mattered more, and this scorecard is built to be re-marked as the evidence lands.

How well has each company kept its roadmap?

Akamai's roadmap used to be the vaguer of the two: a cloud transition described in adjectives. The Q3 2026 call changed the standard. The company now has dated revenue ranges, a published returns frame, secured power, and a named first slip, all in the public record, all gradeable quarter by quarter. A roadmap that can fail in public is worth more than one that can only succeed in narrative.

Cloudflare's roadmap is a compounding promise: keep growing near 30% or better, reach revenue above $5B by the end of 2028, and lift operating margin toward 30% and free cash flow margin toward 30% to 35% over time. The revenue half is tracking ahead of plan. The margin half rests on stock compensation discipline the company has not yet shown, and its one visible software miss, semantic caching, sits exactly in the AI layer where this comparison will be decided.

Akamai commitmentSetStatus at November 7, 2026Grade
Anthropic capacity build, about $5.5B of capex, services from the second half of 2027September 2026On plan with one disclosed slip: $30M of revenue shifted into 2027 on third-party data center punch-list itemsMet so far, slip flagged
2027 CIS revenue of $1.1B to $1.3BNovember 5, 2026Pending; phased 25% to 30% in the first half and 70% to 75% in the second, built on signed deals onlyPending
2027 growth of 17% to 22% in constant currency; 2028 revenue above $7.0BNovember 5, 2026Pending; management says achievable from signed business alone, with no new Anthropic orders assumedPending
Published returns frame: about 13% IRR, about 15% ROIC at full rampNovember 5, 2026Disclosure delivered in full; the returns themselves grade at deliveryPending
Capital discipline: buybacks funded the return story, paused to fund contracted build2026About $616M repurchased, then paused as statedMet
Cloudflare commitmentSetStatus at November 7, 2026Grade
FY2026 revenue guidanceRaised three times during 2026Now $2.963B to $2.964B; Q3 printed up 39% and Q4 is guided up about 38%Beaten to date
Revenue above $5B by the end of 2028Analyst Day, June 2026Tracking ahead: the 2026 guide alone implies a three-year path above the target if growth holds near 30%Pending, ahead
Long-term operating margin of 30% and free cash flow margin of 30% to 35%Analyst Day, June 2026Q3 non-GAAP operating margin of 17.5% and free cash flow margin of 10% leave the margin half unproven; stock compensation near 21% of revenue is the gapPending
GAAP profitabilityGuided on the Q3 call for Q4 2026Pending one quarter; the Q3 GAAP net loss was $34.1MPending
Semantic caching for AI GatewayDescribed as planned since 2024Still unshipped in late 2026 documentation; the shipped cache remains exact-matchMissed to date

What could go wrong for each?

For Akamai, the risk is execution wearing three costumes at once. The company will spend $2.4B to $2.6B on capex in a single quarter against revenue that starts across 2027; a second delay would leave it carrying the depreciation, the debt, and the credibility cost simultaneously.

Anthropic is $11.6B of a $14B book, so one counterparty's health, honesty, and continued preference for CPU capacity underwrite the entire ramp, and the contract's payments are conditional on delivery and service availability, with termination rights for material outage or uncured breach. The returns frame is published precisely so this can be graded, and grading cuts both ways.

For Akamai's base business, the risk is quieter. Security at 55% of revenue funds the transition, and its growth already sits at the low edge of double digits; a slide to mid single digits removes the funder before the ramp arrives. Delivery stabilizing at down 3% is acceptable and even helpful. A return to the steeper declines of earlier years, at the same moment the capex peaks, is the combination the balance sheet least wants. Founder succession, unpriced and undiscussed, sits behind all of it at a company founded in 1998.

For Cloudflare, the risk is the price itself. At 35x trailing sales the stock does not need bad news to fall; it needs ordinary news.

Growth of 39% on a $3B base is genuinely exceptional, and the base triples every three years at that pace, which means the same percentage gets harder in dollars every year. A deceleration into the mid-20s, the normal fate of large growth companies, would reprice the multiple and the narrative in the same week, and the stock has already shown, falling from a $370 high to the $160s within its 52-week range, how quickly that repricing moves.

For Cloudflare's model, the risk is that the growth is being bought at a rising cost. Stock compensation near 21% of revenue means shareholders fund the expansion twice, once in the multiple they pay and once in the dilution they absorb.

Gross margin has slipped about three points in a year as AI inference and large enterprise deals mix in; if agentic traffic monetizes at structurally lower margins than the legacy platform, the 30% operating margin target and the 30% to 35% free cash flow margin target recede together, and those two targets are the entire bridge between today's price and a defensible one.

What is the final score?

SectionAkamaiCloudflare
1. Business diversity78
2. Product portfolio79
3. Edge compute and AI97
4. Network scale98
5. Growth610
6. Profitability and cash flow76
7. Balance sheet and capital allocation76
8. Valuation94
9. Leadership99
10. Moat89
Total7876
  • What moved Akamai from 79 to 78. Growth rose 5 to 6 on contracted acceleration. Profitability fell 8 to 7 and balance sheet fell 8 to 7 on the capex trough, higher leverage, and the paused buyback.
  • What moved Cloudflare from 74 to 76. Profitability rose 5 to 6 on a record non-GAAP operating margin with free cash flow positive through acceleration. Valuation rose 3 to 4 because the multiple compressed from about 47x to 35x trailing sales while growth sped up. Growth was already 10 and stays there.
  • The symmetric standard. Both companies were graded on delivered numbers first and guidance second, with the same evidence bar for a point in either direction.

What are the open gaps and caveats?

  • Cloudflare discloses no product-level revenue. Its mix and portfolio analysis rests on customer counts, net retention, and bookings metrics rather than segment financials.
  • Cloudflare's trailing revenue of $2.73B is settled from its own releases: $2,118.3M for the first nine months of 2026 plus about $614M for Q4 2025. Aggregator tables still showing $2.17B are displaying fiscal 2025 revenue, not the trailing year. One aggregator's trailing free cash flow figure of about $626M could not be reproduced from the releases; the release-derived figure is about $322M.
  • Akamai's debt is as of June 30, 2026, the latest filing available at rebuild; its cash balance is September 30, 2026. Akamai's trailing revenue of about $4.4B is derived from quarterly prints and the FY2026 guide.
  • Cloudflare's 2027 revenue estimate near $3.7B predates the Q3 print (CFRA and S&P Global). Post-print revisions will likely raise it, which would lower the 2027 forward multiple modestly.
  • Neither company publishes cache hit rates, per-location GPU counts, or cloud segment margins. The AI sections rest on announced deployments, contract disclosures, and product documentation.
  • Akamai's book economics (about $2.2B of revenue at full ramp, about 75% adjusted EBITDA margin, about 13% after-tax IRR, about 15% adjusted ROIC) and Cloudflare's long-term targets (revenue above $5B by 2028, 30% operating margin, 30% to 35% free cash flow margin) are management statements, labeled as such wherever they are used.
  • Network capacity figures come from mixed primary and secondary sources and are not audited.

Sources and basis

  • Akamai Q3 2026 earnings release and the full earnings call, November 5, 2026, graded in our Q3 2026 earnings analysis.
  • Cloudflare Q3 2026 earnings release via Business Wire, November 5, 2026, including the condensed consolidated statements and reconciliations.
  • Cloudflare Analyst Day materials, June 2026, for the 2028 revenue target and long-term margin targets.
  • Prices as of the November 6, 2026 close via Finnhub. Share counts from the Q3 releases.
  • Background research: our Akamai five-pillar analysis and the AI Grid deep dive.

Disclosures. This comparison is published for educational purposes only and is not investment advice. The author is long Akamai Technologies (AKAM) and is a former Akamai employee (2013 to 2018); nothing here draws on non-public information from that employment. The author holds no position in Cloudflare (NET). See the full disclosures page.