API Gateway Pricing: Models, Cost Drivers, and Examples
April 15, 2024
API gateway pricing depends on more than the sticker price. The final API gateway price can change based on deployment model, request volume, data transfer, plugin requirements, support, observability, high availability, and the engineering time required to operate the gateway.
For FinOps practitioners, platform teams, and architects, the goal is not simply to find the lowest monthly fee. It is to choose an API gateway and API management model that fits traffic scale, security requirements, cloud strategy, and operational maturity. This article explains common pricing models, compares cost drivers across managed and self-managed options, and shows why open-source gateways such as Apache APISIX and enterprise platforms such as API7 Enterprise should be evaluated through total cost of ownership.
Pricing changes frequently. Use this article to understand pricing logic and cost drivers, then confirm current numbers on each vendor's official pricing page before budgeting.
Understanding API Gateway Pricing Models
Before examining each provider's specifics, let's establish a foundational understanding of API gateway pricing models. Typically, these models revolve around factors directly tied to API usage. Key concepts include:
-
Pay-per-Use vs. Subscription: While some API gateways operate on a pay-per-use model like Amazon API Gateway, others, such as Apigee, employ subscription models with tiered functionalities and request volumes.
-
Throttling Limits: To maintain operational stability and prevent misuse, most providers implement throttling limits, restricting the volume of requests an API can process per second or within a designated timeframe. Breaching these thresholds could result in supplementary charges.
-
Data Transfer Costs: Keep a keen eye on data transfer expenses, particularly in cloud-based solutions, as they could be incurred based on the volume of data flowing through the API gateway.
A firm grasp of these fundamental principles will empower you to thoroughly assess the nuanced pricing models presented by each API gateway provider in the subsequent sections. Moreover, it's imperative to factor in the Total Cost of Ownership (TCO) when deliberating your choices. TCO extends beyond the initial price tag to encompass support expenses, potential resource consumption (especially for self-managed deployments), and any concealed charges linked with the service.
| Pricing model | Common in | What to watch |
|---|---|---|
| Pay per request | Cloud-managed gateways | Request spikes, retries, regional pricing, data transfer |
| Subscription tier | Enterprise API management platforms | Feature limits, support tier, environments, traffic entitlements |
| Open source self-managed | Apache APISIX, Kong OSS, other OSS gateways | Infrastructure, SRE time, upgrades, observability, support |
| Hybrid control plane | Enterprise or cloud control plane with self-managed data plane | Data-plane infrastructure plus platform subscription |
| Custom enterprise pricing | Large-scale or regulated deployments | SLA, support scope, compliance requirements, dedicated environments |
Deep Dive into Individual Pricing Structures
After thoroughly understanding the basic framework of API gateway pricing, let's delve into the specific pricing details of each selected vendor.
Amazon API Gateway
-
Pricing Model: Pay-per-use. You're charged based on the number of API requests received and the amount of data transferred out.
-
Free Tier: Amazon API Gateway provides a versatile pricing structure comprising three tiers: Free, Standard, and Usage. The Free tier grants access to one million API calls monthly for up to 12 months, thereafter transitioning to a usage-based billing model determined by API call volume and data transfer. Starting at $1.00 per million API calls, the Standard tier extends advanced functionalities such as custom domain integration, SSL/TLS certification, and Cross-Origin Resource Sharing (CORS) support. The Usage tier caters to high-traffic APIs with tiered pricing commencing at $0.90 per million API calls.
-
Key Cost Factors:
-
Number of Requests: The primary cost driver is the volume of API requests you receive. Costs increase incrementally based on tiers (e.g., first 1 million requests free, next tier priced per million requests).
-
Data Transfer Out: Data transferred out of the gateway incurs charges, but there are no charges for private APIs. Consider integrating with other AWS services for potential cost savings.
-
Throttling Limits: Free tier limitations apply (throttling limits may be lower than in paid plans). Upgrading to paid plans offers increased throttling limits for handling higher traffic volumes.
-
-
Additional Considerations:
-
Reserved Instances: For predictable workloads, reserving instances can offer significant cost savings compared to on-demand pricing.
-
Integration with AWS Services: Utilizing other AWS services alongside API Gateway can lead to cost optimizations through bundled pricing or service-specific discounts.
-
Cost of Vendor Lock-in: If you need to consider deploying API gateways in multi-cloud and hybrid clouds in the future, then the costs of vendor lock-in and migration need to be considered. For infrastructure like API Gateway, we certainly don't want frequent changes and migrations.
-
Apigee
-
Pricing Model: The service operates on a subscription model with multiple tiers (Standard, Enterprise, Enterprise Plus), each tailored to specific needs and offering distinct features alongside allocated API request volumes. For instance, the Standard tier encompasses 1.25 billion standard API proxy calls and 250 million extensible API calls. Pricing stands at $20 per million standard API proxy calls and $100 per million extensible API calls. It is typically positioned as a broader enterprise API management platform rather than only a lightweight gateway.
-
Free Tier: A limited free tier may be available, but it typically includes restricted functionality and a low request quota.
-
Key Cost Factors:
-
Subscription Tier: The chosen subscription tier significantly impacts cost. Higher tiers offer more features, increased request volumes, and potentially advanced security or analytics capabilities (which may incur additional charges).
-
Additional Features: For instance, API Analytics costs $20 per 1 million API calls, while Advanced API Security is priced at $350 per 1 million API calls.
-
Traffic Volume: If your usage surpasses the allocated request quota within your selected tier, you will incur additional charges based on usage.
-
Apigee Edge: For on-premises deployments, Apigee Edge has a separate pricing structure that may involve upfront licensing fees or ongoing subscription costs.
-
Cost of Vendor Lock-in: The price is akin to that experienced with Amazon API Gateway.
-
Kong
-
Pricing Model: The open-source version is freely downloadable and deployable, but it demands internal expertise for effective management and maintenance. Kong Konnect, on the other hand, represents Kong's cloud iteration, featuring a pricing model akin to Apigee's, offering a tiered structure coupled with a pay-as-you-go approach. Primarily, Kong Konnect levies charges for services and API calls, priced at $105 per service per month and $34.25 per million requests respectively. Notably, if the number of services exceeds 100, the overall cost surpasses that of Apigee.
-
Free Tier: The open-source Kong is a free tier, offering core API gateway functionality.
-
Key Cost Factors:
-
Enterprise Edition: The enterprise edition offers exclusive plugins, such as OIDC, unavailable in the open-source version, along with comprehensive technical support services.
-
Self-Managed vs. Cloud-Based: While the core software is free, self-managing Kong requires additional resources such as servers and technical expertise, potentially impacting cloud expenditure. Managed cloud deployments of Kong may also be available with distinct pricing structures.
-
Complex Pricing Model: Taking the Konnect Plus tier as an example, the pricing model encompasses multiple dimensions, including Gateway Services, API Requests, Paid Plugins, Premium Plugins, API Analytics, and Zones. Estimating expenses and understanding billing will pose considerable challenges.
-
APISIX
-
Pricing Model: APISIX stands as a top project within the Apache Software Foundation. API7.ai, the donor of APISIX, offers both enterprise and Cloud versions built upon Apache APISIX. The pricing model for API7 Enterprise is structured around CPU counts, with subscription fees independent of API call volume—a notable departure from Kong's Enterprise pricing paradigm.
-
Free Tier: The open-source APISIX offers core API gateway features at no cost.
-
Key Cost Factors:
-
Enterprise Edition: Enhanced functionalities such as high availability clustering, gateway groups, and SLA support necessitate additional enterprise licenses.
-
Deployment Options: Analogous to Kong, self-managed deployments demand supplementary resources, potentially influencing costs. Moreover, API7 Cloud exclusively manages the control plane, while the data plane mandates provisioning your own machine resources, thereby ensuring data compliance and security.
-
By understanding these individual pricing structures and key cost factors, you can make a more informed decision when choosing the right API gateway for your budget and specific needs.
Practical API Gateway Cost Scenarios
The right pricing model often depends on the operating scenario.
| Scenario | Pricing risk | Better fit |
|---|---|---|
| Early-stage API or low traffic | Engineering time can exceed software cost | Managed gateway or simple self-managed APISIX deployment |
| High-volume public API | Per-request billing can grow quickly | Open-source or enterprise gateway with predictable capacity-based pricing |
| Regulated enterprise API program | Compliance, audit logs, support, and access controls matter | Full API management platform such as API7 Enterprise |
| Multi-cloud or hybrid architecture | Cloud-native gateways can create lock-in | Portable gateway such as Apache APISIX with centralized management |
| AI or LLM traffic | Token cost, retries, streaming, and fallback affect spend | AI Gateway pattern with token-aware routing and observability |
FAQ
How is API gateway pricing calculated?
API gateway pricing is commonly calculated by API calls, data transfer, subscription tier, number of services or routes, deployment environments, support level, or infrastructure usage. Self-managed open-source gateways may have no license fee, but still require infrastructure and operations work.
Why can API gateway price increase at scale?
Costs can rise when traffic grows, retries increase request volume, data transfer crosses tiers, advanced plugins require paid plans, or teams need higher availability and support. Always model both average and peak traffic.
Is an open-source API gateway free?
Open-source gateways such as Apache APISIX are free to download and use, but production deployments still have infrastructure, monitoring, security, and maintenance costs. Some organizations add enterprise support or a management platform for governance and operational efficiency.
What should I compare before choosing an API gateway?
Compare runtime capabilities, deployment model, security controls, observability, support, operational effort, and total cost. The API gateway software comparison and API management platform guide can help structure the evaluation.