Cloudflare Workers vs AWS Lambda: Costs by Workload
Compare Cloudflare Workers and AWS Lambda pricing with explicit workload estimates, CPU versus duration billing, and the costs excluded from headline rates.

Cloudflare Workers can suit requests that spend time waiting for other services, while AWS Lambda can cost less for small workloads covered by its monthly allowances. For buyers comparing Cloudflare Workers vs AWS Lambda pricing, the decisive serverless computing costs depend on active CPU time, allocated memory and measured execution duration.
- Workers Paid starts at $5 monthly, including 10 million requests and 30 million CPU milliseconds.
- Lambda Event Functions include one million requests and 400,000 GB-seconds monthly; eligibility and remaining allowances matter.
- Workers excludes network waiting from CPU time; Lambda Event Functions use memory-weighted execution duration.
- The estimates below cover function charges, with explicit assumptions—not a complete application invoice.
What Each Platform Actually Bills
Data checked: November 16, 2026.** Prices are USD. Lambda estimates use US East, Northern Virginia, first-tier x86 on-demand pricing. Workers uses its published Standard usage model. These are purchasing references for US, UK, Canadian, Australian and Singaporean readers, rather than local tax-inclusive quotations.
The Cloudflare Workers Paid plan includes the allowances above, then charges $0.30 per additional million requests and $0.02 per additional million CPU milliseconds. Cloudflare lists no additional Workers bandwidth or egress charges. Database and other product usage still needs its own assessment.
AWS Lambda pricing lists $0.20 per million requests. Its Northern Virginia examples use $0.0000166667 per GB-second for x86 execution. Monthly allowances must be applied across applicable usage, rather than assigned afresh to every function.
Scope matters: AWS documentation now distinguishes Event Functions, Web Functions and MicroVMs. This guide models the original event-driven function type, now called Event Functions. Its duration-based arithmetic should not be applied to the other products.
That distinction also changes how buyers should read edge computing platform pricing. A request count alone leaves out the resource being purchased. Start by identifying the execution model, then measure the application against its billing units. A low request rate cannot establish the lower total bill when compute consumption differs.
Explicit Workloads Show Why the Answer Changes
The following are editorial calculations using the official rates, not performance benchmarks. Each invocation is assumed to count as one billable request. Workers consumes 7 milliseconds of CPU per invocation; Lambda has 1,024 MB allocated memory and the billed durations specified below. These measurements are hypothetical inputs, not claims that identical code achieves them.
All monthly allowances are assumed available. Exclusions include taxes, credits, storage, logging, external services, networking charges outside the quoted Workers terms, and optional features. Amounts are rounded to cents.
| Platform and scenario | Assumed monthly workload | Function charges, USD | Potential advantage and buyer profile |
|---|---|---|---|
| Workers Paid, small service | 1 million requests; 7 ms CPU each | $5.00 | Included capacity; buyer accepting a subscription floor |
| Lambda Event Functions, small service | 1 million requests; 1,024 MB; 100 ms billed duration | $0.00 | Available allowances cover this usage; small-service buyer |
| Workers Paid, waiting-heavy service | 15 million requests; 7 ms CPU each | $8.00 | CPU meter excludes network waits; lightweight request handling |
| Lambda Event Functions, waiting-heavy service | 15 million requests; 1,024 MB; 500 ms billed duration | $121.13 | Configurable memory; buyer needing the event-function model |
These rows compare assumed demand and metered configurations, rather than equivalent processing capacity. Cloudflare's $8 calculation also appears in its official pricing example. The Lambda estimate applies its published request and GB-second rates after allowances.
Before using an AWS Lambda cost calculator, replace these inputs with measured billing data. Keep function execution separate from upstream service spending. An application calling a model should also budget that service independently; see AI Inference Pricing: AWS, Azure and Google Cloud Costs.
Memory, Waiting and Engineering Change the Decision
Cloudflare defines CPU time as time spent executing code. Waiting for network requests, database queries or KV reads does not count toward that meter. This makes the CPU-to-elapsed-time relationship a useful purchasing input for an application that calls other services.
Workers also has a documented 128 MB memory limit per isolate, which may handle multiple concurrent requests. That is a capacity constraint, not a selectable memory tier. Test concurrent traffic and large responses before assuming that a low calculated bill makes the runtime suitable.
AWS memory documentation allows Event Functions to configure 128 MB to 10,240 MB, with CPU allocated proportionally. Increasing memory can improve execution performance. Consequently, selecting the smallest allocation does not establish the lowest execution cost: buyers need to measure how duration changes with allocation.
A useful trial should therefore record successful requests, errors, memory consumption, execution time and the associated bill. Include initialization and extension effects in the measured billed duration rather than substituting a stopwatch around only the business logic. Reproduce the expected dependency calls and response sizes.
The FinOps Foundation's Usage Optimization framework recommends weighing optimization benefits against engineering effort, risk and disruption. Apply that principle to a migration: price the work required to change dependencies, deploy safely and operate the replacement.
Keep procurement categories separate as well. Endpoint Security Pricing 2026 addresses another business technology budget; its costs should not be folded into a serverless execution comparison without an explicit allocation method.
Choose From Measured Requirements
Choose Workers for evaluation when requests perform modest computation, spend substantial time waiting on dependencies and fit its memory constraints. Choose Lambda Event Functions for evaluation when configurable memory or an existing event-driven AWS design is central to the application. Neither recommendation establishes a universal price winner.
For an existing production service, compare incremental costs with the allowances actually remaining in the account. For a new service, model both quiet months and expected production demand. Require the estimate to identify its region, architecture, execution model, dependency charges and excluded services.
The purchasing decision should follow a successful workload test and a complete cost worksheet. A cheaper execution estimate is useful only when the platform also meets the application's capacity, reliability and operating requirements.
Take the quizWhat Kind of Holiday Spender Are You? Take the Black Friday QuizStart →- 1I’m A Celebrity 2026 Line-Up: Rooney Leads 11 Stars Into Jungle

- 2Mavericks Surfers Race to California for Waves Near 50 Feet

- 3NYC Medical Examiner Has Not Determined Cause of Death of Beau Rose Jessup

- 4Mountaineer Finds Ice Age Cave Bear Graveyard in the Dolomites

- 5Rugby League World Cup Final 2026: Kick-Off Times and How to Watch Kangaroos v New Zealand

Lifestyle7 Signs Your Social Battery Is Running Low (and Why You Keep Cancelling Plans)
Lifestyle · Food & HealthThanksgiving Food Safety: How Long Leftovers Last and the Safe Turkey Temperature
Lifestyle7 Quiet Habits of Emotionally Intelligent People, According to Psychologists
LifestylePeople Who Don’t Like Celebrating Their Birthday Often Share These 7 Traits















