API Gateway AWS Calculator
Estimate monthly Amazon API Gateway costs for HTTP APIs, REST APIs, and WebSocket APIs using traffic volume, payload size, connection minutes, and internet egress assumptions. This calculator is designed for fast budget planning, architecture reviews, and early cost modeling.
Pricing hint: HTTP APIs are estimated at $1.00 per million for the first 300 million requests, then $0.90 per million. Manual data transfer out overrides the automatically estimated traffic volume when greater than zero.
How to Use an API Gateway AWS Calculator for Accurate Cloud Cost Planning
An API Gateway AWS calculator helps engineering leaders, DevOps teams, product managers, and finance stakeholders estimate one of the most common costs inside a modern serverless or microservices stack: the price of exposing APIs to internal services, web applications, mobile apps, partners, and real time clients. Amazon API Gateway looks simple at first because many teams think in terms of “requests per month,” but billing can vary significantly depending on API type, request volume, message size, connection duration, and internet egress.
If you want a realistic estimate, you need to separate three things. First, identify whether you are using HTTP APIs, REST APIs, or WebSocket APIs. Second, estimate your monthly traffic profile, not just average daily traffic. Third, decide whether internet data transfer out should be modeled separately, because network egress is often meaningful at scale. This page is built to do exactly that.
At a strategic level, API Gateway pricing matters because it directly affects architecture choices. A startup launching a public mobile backend may prioritize speed and choose HTTP APIs for lower request pricing. An enterprise with heavy transformation rules, request validation, usage plans, or older integrations may keep REST APIs despite higher request costs. A collaboration product that streams events to browsers may need WebSocket APIs, where message units and connection minutes become the real pricing drivers.
What This Calculator Estimates
This calculator focuses on the core variables most teams need during planning:
- HTTP API request pricing, which is generally the lowest cost option in Amazon API Gateway for standard request and response patterns.
- REST API request pricing, which is usually higher but can align better with legacy patterns or advanced API Gateway features.
- WebSocket pricing, which includes both message charges and connection minute charges.
- Estimated or manual internet egress, because data transfer out can become material with large payloads or global user bases.
Important planning note: API Gateway request costs are only one part of the total architecture. In production, you should also model AWS Lambda duration, Lambda requests, VPC charges if applicable, CloudWatch logging, WAF, DynamoDB or Aurora usage, and observability tooling.
Understanding the Main API Gateway Billing Models
HTTP APIs
HTTP APIs are usually the lowest cost choice for modern stateless APIs. They are commonly used with Lambda, containers, and HTTP backends. For many workloads, the lower per million request price makes HTTP APIs attractive when traffic grows quickly. Teams often choose them for public mobile applications, single page app backends, and serverless microservices that do not need every advanced REST API feature.
REST APIs
REST APIs generally carry a higher per request price, but they may still be the right fit when an organization depends on long established API Gateway features, existing deployment pipelines, or integration patterns that are already standardized. In enterprise portfolios, the higher unit price is often accepted because the migration cost to another API shape may exceed the monthly savings.
WebSocket APIs
WebSocket APIs are different. Billing is based on both messages and connection minutes. This makes them ideal for chat, dashboards, collaborative tools, IoT control channels, and event driven interfaces, but they require more careful forecasting. A product with modest message volume and very long connection times may behave differently from one with heavy bursts of short lived messages.
| API type | Typical billing unit | Representative published pricing statistic | Best fit |
|---|---|---|---|
| HTTP API | Requests, per million | About $1.00 per million requests for the first 300 million monthly, then about $0.90 per million after that | Cost efficient public or internal APIs |
| REST API | Requests, per million | About $3.50 per million requests for the first 333 million monthly, then about $2.80 per million after that | Legacy compatibility and advanced gateway features |
| WebSocket API | Messages and connection minutes | About $1.00 per million messages plus about $0.25 per million connection minutes | Real time and bidirectional communication |
The numbers above are widely cited for common AWS pricing examples and are very useful for planning, but you should always validate against the current AWS regional pricing page before committing budget. Rates can vary by region and product updates.
Why Payload Size Still Matters
Many teams assume payload size is irrelevant because request pricing is request based, not byte based, for HTTP APIs and REST APIs. That is only partly true. Payload size matters for two reasons. First, larger payloads increase network egress. Second, in WebSocket workloads, message metering is based on 32 KB message units. If your application sends 33 KB per message instead of 32 KB, the message can effectively count as two billed units rather than one. That single design detail can materially alter your monthly bill.
This is why an API Gateway AWS calculator should not stop at pure request count. It should also let you model payload behavior, especially if your service returns large JSON responses, image metadata, analytical records, or frequent event updates.
| Average WebSocket payload | Billed message units | Effective multiplier | Impact on 10 million app messages |
|---|---|---|---|
| 1 KB | 1 unit | 1x | 10 million billed units |
| 32 KB | 1 unit | 1x | 10 million billed units |
| 33 KB | 2 units | 2x | 20 million billed units |
| 65 KB | 3 units | 3x | 30 million billed units |
Step by Step: How to Estimate API Gateway Cost Correctly
- Choose the API type. Start with the actual deployment choice, not the one you hope to migrate to later.
- Estimate monthly request or message volume. Use peak month planning if you are budgeting for production capacity.
- Enter the average payload size. This is especially valuable for egress estimation and WebSocket metering.
- Add monthly connection minutes for WebSocket APIs. If clients stay online for long periods, connection minutes can be substantial.
- Set the internet egress rate. The default planning rate in this calculator is $0.09 per GB, but your actual number may differ by region and architecture.
- Use manual data transfer when you have a stronger network estimate. This is useful when payload size does not reflect compression, caching, or mixed response sizes.
Examples of Real World Cost Scenarios
Scenario 1: Public HTTP API for a Mobile App
Suppose a mobile application generates 50 million API calls per month, with an average payload size of 8 KB. HTTP API pricing will usually be the lowest request cost option. If the service is read heavy and responses are small, the request charge may remain a manageable line item. In this case, developers should compare API Gateway cost with downstream Lambda, auth service calls, and database reads. Often API Gateway is not the dominant cost, but when traffic reaches hundreds of millions of requests per month, it deserves direct optimization.
Scenario 2: Legacy REST API with Enterprise Controls
An enterprise team may run 120 million monthly requests through REST APIs because they rely on mature policy patterns or already have established tooling. The price per million requests is higher than HTTP APIs, but the migration cost could involve security reviews, regression testing, and partner onboarding. In that situation, a calculator helps answer a better question than “Which unit price is lowest?” The better question is “What is the break even point for migration?”
Scenario 3: Real Time Dashboard with WebSocket APIs
Consider a SaaS dashboard with 2 million monthly connection minutes and 15 million messages. If each message is only 2 KB, message cost remains close to the nominal per million rate. If product changes push each message to 40 KB, billed message units roughly double. That means a seemingly small product design decision can create a meaningful increase in operating expense.
How to Reduce API Gateway Spend Without Hurting Reliability
- Use HTTP APIs where feature parity is sufficient. This is often the cleanest way to reduce per request cost.
- Compress payloads and trim response fields. Smaller payloads reduce egress and can improve client performance.
- Cache at the client, edge, or application layer. Fewer repeat calls means fewer billed requests.
- Batch updates for WebSocket workloads. Sending fewer, more meaningful messages can lower billed message volume.
- Watch for accidental polling. Poor client retry logic can drive request counts far above expected levels.
- Instrument request distribution by endpoint. One noisy endpoint can account for a disproportionate share of spend.
Architecture, Governance, and Security Considerations
Cost calculators are most useful when they sit inside a broader governance process. API management decisions involve security controls, service isolation, traffic governance, and reliability engineering. Organizations that handle public sector, education, healthcare, or financial workloads often need to evaluate API exposure under formal security guidance. The following public resources are helpful for architectural context:
- NIST: The NIST Definition of Cloud Computing
- NIST: Building Secure Microservices-Based Applications
- CISA: Secure by Design
These sources are not pricing references, but they are relevant to the decisions that influence API gateway design, especially in regulated environments where security architecture can shape service boundaries, gateway usage patterns, and operational overhead.
Common Mistakes When Estimating API Gateway Bills
- Ignoring monthly peaks. Average daily traffic often understates launch events, batch spikes, or partner sync windows.
- Forgetting egress. Teams may focus on request pricing while underestimating network transfer out.
- Modeling WebSocket traffic as simple requests. Connection minutes and message unit rounding must be included.
- Excluding retries and client errors. Retries count toward billable traffic and can be surprisingly expensive.
- Assuming one region price fits all. Validate the regional pricing page before finalizing forecasts.
- Not comparing API Gateway cost with engineering cost. A lower monthly service charge may not justify a disruptive migration.
Final Takeaway
An API Gateway AWS calculator is most valuable when it helps you make decisions, not just produce a number. The right estimate combines API type, traffic volume, payload behavior, connection duration, and egress assumptions into one view. For small deployments, the difference between HTTP API and REST API may be modest in absolute dollars. At enterprise or consumer scale, that same difference can become significant enough to influence product design, migration timing, and platform standards.
Use the calculator above to build a fast estimate, then validate the result against your architecture, AWS regional pricing, and traffic telemetry. When used this way, an API Gateway calculator becomes a practical planning instrument for cloud budgeting, capacity reviews, and technical strategy.