New See exactly what you're overpaying AWS in under 60 seconds. Estimate your AWS savings →

GCP Load Balancer Pricing (2026): Rates, Costs & Examples

GCP Load Balancer pricing starts at $0.025/hour total for up to five forwarding rules, or approximately $18.25/month. In Iowa (us-central1), external load balancers also charge $0.008/GiB for inbound and outbound data processing. Internal Application Load Balancers have a minimum proxy charge of 0.075/hour(54.75/month), even without traffic.
Picture of Haider Ali
Cloud & FinOps Content Specialist

Haider Ali

Cloud & FinOps Content Specialist
November 13, 2026

·

19 min read
GCP Load Balancer Pricing (2026): Rates, Costs & Examples
In this article
Key takeaways
1
Forwarding rules: Up to five global forwarding rules cost $0.025/hour total per project. Regional rules have a separate pricing group.
2
Traffic processing: External load balancers in Iowa charge $0.008/GiB for inbound and outbound processed data.
3
Idle costs: Forwarding rules remain billable without traffic. Internal Application Load Balancers also have minimum proxy charges.
4
Additional costs: Internet egress, backend compute, Cloud CDN, Cloud Armor, and logging can increase the bill.

GCP Load Balancer Pricing by Type

Google Cloud offers Application Load Balancers for HTTP(S) traffic and Network Load Balancers for TCP, UDP, and other supported protocols. Pricing differs based on whether traffic is public or private and whether the configuration requires managed proxy capacity.

The table below compares common configurations using published USD rates for Iowa (us-central1).
Load balancer type Fixed monthly cost Data processing
External Application LB $18.25 for up to 5 global rules $0.008/GiB inbound and outbound
External Network LB $18.25 for up to 5 rules in one pricing group $0.008/GiB inbound and outbound
Internal Application LB $54.75 minimum proxy cost per forwarding rule $0.008/GiB
Internal passthrough Network LB $18.25 for up to 5 regional rules $0.008/GiB inbound and outbound
Prices assume 730 hours and the listed configurations. Network Load Balancer subtypes can have different billing behavior. Internal Application Load Balancers use a separate proxy-based pricing model. Rates vary by region.

Source: Google Cloud Load Balancing pricing.

External Application Load Balancer

Best for: Public websites, HTTPS applications, and APIs.

External Application Load Balancers operate at Layer 7, allowing traffic routing based on domains, URL paths, and HTTP requests. Global configurations can distribute traffic across regions, while regional configurations serve applications from a selected region.

Both global and regional configurations have applicable forwarding-rule and data-processing charges. Regional external Application Load Balancers support Premium and Standard Network Service Tiers, while global external Application Load Balancers use Premium Tier.

If your application runs on virtual machines, our Google Compute Engine pricing guide explains the backend compute costs.

External and Internal Network Load Balancers

Best for: TCP, UDP, and other network-level workloads.

Network Load Balancers operate at Layer 4. Google Cloud offers passthrough configurations, which forward traffic to backends without terminating the connection, and proxy configurations, which terminate client connections.

For example, an internal passthrough Network Load Balancer with one regional forwarding rule running for 730 hours incurs approximately $18.25 in forwarding-rule charges. Processing 500 GiB at $0.008/GiB adds $4, bringing the load-balancing estimate to $22.25/month.

Proxy Network Load Balancers have different architectural and capacity considerations, so confirm the selected subtype before estimating costs.

Internal Application Load Balancer

Best for: Private APIs, microservices, and internal HTTP(S) applications.

Internal Application Load Balancers use managed Envoy proxy instances. Google Cloud allocates at least three proxies per internal managed forwarding rule, including when traffic is zero.

In Iowa, the minimum monthly proxy cost is:
3 × $0.025 × 730 = $54.75
Additional proxies are billed when bandwidth, connections, or request volume require more capacity.

How GCP Load Balancer Billing Works

For most external configurations, the basic calculation is:
Monthly load-balancing cost = Forwarding-rule charges + Data-processing charges
Internal Application Load Balancers instead use proxy-instance and processing charges.

1. Forwarding-Rule Charges

A forwarding rule specifies the IP address, protocol, and port through which traffic reaches a load balancer.

Google Cloud charges $0.025/hour total for the first five global forwarding rules within a project. Each additional rule costs $0.01/hour.
Global forwarding rules Hourly cost Monthly cost
1 $0.025 $18.25
3 $0.025 $18.25
5 $0.025 $18.25
6 $0.035 $25.55
10 $0.075 $54.75
Global and regional forwarding rules have separate pricing groups. For example, one global rule and one regional rule can produce two separate $0.025/hour charges.

2. Data-Processing Charges

In Iowa, external load balancers charge $0.008/GiB for inbound and outbound processed traffic.

The formula is:
Processing cost = (Inbound GiB + Outbound GiB) × $0.008
If a load balancer processes 500 GiB inbound and 1,000 GiB outbound, the processing charge is $12. Add the applicable forwarding-rule charge to calculate the load-balancing total.

3. Does an Idle GCP Load Balancer Cost Money?

Yes. Forwarding rules remain billable even when they receive no requests. An external configuration with one forwarding rule can therefore cost $18.25/month while idle.

Internal Application Load Balancers also retain their minimum proxy allocation, as explained above. Stopping backend virtual machines does not automatically remove these charges.

GCP Load Balancer Monthly Cost Examples

The following examples use Iowa pricing and 730 hours per month. They exclude internet egress, backend compute, and optional services.

Example 1: Small Website

A small business runs a public website behind one external Application Load Balancer. Monthly traffic includes 100 GiB inbound and 500 GiB outbound.
Component Calculation Monthly cost
Forwarding rule $0.025 × 730 $18.25
Inbound processing 100 × $0.008 $0.80
Outbound processing 500 × $0.008 $4.00
Total $23.05
The website’s estimated monthly load-balancing cost is $23.05.

Example 2: High-Traffic Production Application

A production application uses a global external Application Load Balancer. It processes 2,048 GiB inbound and 4,096 GiB outbound per month.
Component Monthly cost
Forwarding rule $18.25
Inbound processing $16.38
Outbound processing $32.77
Total $67.40
The unrounded calculation is:
$18.25 + (6,144 × $0.008) = $67.402
For Kubernetes-based applications, our GKE cost optimization guide covers the cluster and compute expenses behind the load balancer.

Example 3: Internal Application Load Balancer

A company deploys an internal Application Load Balancer for private microservices. The configuration requires only the minimum three proxy instances and processes 500 GiB per month.
Component Monthly cost
3 proxy instances $54.75
500 GiB processing $4.00
Total $58.75
The estimated monthly cost is $58.75. Higher traffic may require additional proxies, increasing the bill.

Hidden GCP Load Balancer Costs

Load-balancing charges are only part of the overall application bill. Depending on your architecture, additional networking, security, and compute services may contribute significant costs.
Additional cost What to check
Internet egress Traffic transferred to internet users
Backend compute Compute Engine, GKE, or Cloud Run usage
Cloud CDN Cache egress, cache fill, and requests
Cloud Armor Security policies and applicable requests
Cloud Logging Log ingestion and storage
Cloud NAT Networking charges for applicable configurations
Cross-region traffic Data transferred between regions
Authorization policies Request, rule, and policy charges

Internet Egress vs. Load-Balancer Processing

These are separate charges. Processing measures traffic handled by the load balancer, while internet egress covers data transferred from Google Cloud to external destinations.

For example, processing 1,000 GiB of outbound traffic at $0.008/GiB costs $8. Internet data transfer is calculated separately using the applicable network pricing.

For broader networking expenses, see our Google Cloud cost management guide.

Cloud Run and GKE Backends

Applications using Cloud Run or GKE generate backend costs in addition to load-balancing charges. Cloud Run usage depends on requests, compute allocation, and scaling, while GKE costs include applicable cluster and node resources.

When an external Application Load Balancer uses a serverless NEG with Cloud Run, the separate Cloud Run outbound data transfer charge does not apply to traffic routed through that integration, although applicable internet egress remains.

See our Cloud Run pricing guide for a detailed backend cost breakdown.

Authorization Policy Pricing

Google Cloud also charges for authorization policies when these features are configured. These policies control whether incoming requests can access an application.
Authorization component External ALB Internal ALB
Request processing $0.75/million requests $0.60/million requests
Authorization rules $1/rule/month $1/rule/month
Authorization policies $5/policy/month $5/policy/month
Certain request-processing charges may be included with qualifying Cloud Armor configurations. Policy and rule charges remain separate.

These are optional authorization-related charges, not fees automatically applied to every load balancer.

Source: Google Cloud Load Balancing pricing.

Why Internal Load Balancers Can Become Expensive

Internal Application Load Balancers scale managed proxy capacity according to traffic requirements. Google Cloud evaluates bandwidth, connections, and requests over ten-minute periods.

According to Google’s published capacity model, one proxy instance supports up to:

18 MB/second of bandwidth

600 new HTTP or 150 new HTTPS connections/second

3,000 active connections

1,400 requests/second with Cloud Logging disabled

Google calculates proxy requirements from bandwidth and connection/request demand, then uses the greater requirement, subject to the minimum allocation.

How Logging Affects Proxy Costs

Cloud Logging can reduce the request-handling capacity of each proxy.

Google states that logging 100% of requests reduces capacity from approximately 1,400 to 700 requests/second per proxy.

Consider an internal application processing 2,800 requests/second:
Configuration Request-based proxy requirement
Logging disabled 2 proxies
100% request logging 4 proxies
The load balancer still requires at least three proxies. Therefore, if bandwidth and connection requirements are low enough, the first configuration uses the three-proxy minimum while the second requires four.

At $0.025/proxy-hour, that difference represents approximately $18.25/month for a continuously sustained workload.

Actual charges depend on the combined capacity calculation, including bandwidth, new connections, and active connections. Logging does not automatically increase every internal load balancer’s bill.

Source: Google Cloud proxy instance pricing and capacity.

GCP Load Balancer Cost Optimization: 7 Ways to Save

1. Remove Unused Forwarding Rules

Forwarding rules can generate charges even without traffic. Review old testing environments, abandoned applications, and unused frontends, then remove resources that no longer support active workloads.

Verify dependencies before deleting a rule, especially when multiple services share a load balancer.

2. Consolidate Forwarding Rules

For HTTP(S) workloads, host-based and path-based routing can allow multiple applications to share one frontend. This may reduce unnecessary forwarding rules and simplify management.

Consolidation is most useful when it eliminates billable rules beyond the included group. Keep applications separate when security or reliability requirements demand it.

3. Use Cloud CDN for Cacheable Traffic

Cloud CDN can serve cached content from Google’s edge network, reducing traffic processed by the load balancer and backend.

However, Cloud CDN introduces its own cache egress, cache fill, and request charges. Compare total costs and cache hit rates before enabling it solely to save money.

4. Block Unwanted Traffic with Cloud Armor

Cloud Armor can block bots and malicious requests before they reach application backends. This may reduce unnecessary compute usage and applicable outbound processing charges.

Cloud Armor also has policy and request-related costs, so evaluate the savings against the additional security expense.

5. Choose the Right Network Service Tier

Regional external Application Load Balancers support both Premium and Standard Network Service Tiers. Standard Tier may offer lower internet transfer costs for suitable applications, while Premium Tier uses Google’s global backbone for performance-focused connectivity.

Compare latency, availability requirements, and network pricing before switching tiers. Global external Application Load Balancers require Premium Tier.

6. Optimize Internal Proxy Usage

Monitor proxy capacity and logging settings using the sizing model explained above. Adjust logging sample rates where appropriate without sacrificing essential observability or security requirements.

7. Optimize Backend Compute

For many applications, backend compute represents a larger expense than load balancing. Rightsize Compute Engine instances, improve GKE node utilization, and evaluate eligible compute commitments.

Our GCP cost optimization best practices guide covers additional ways to reduce infrastructure waste.

Do GCP Committed Use Discounts Cover Load Balancers?

No, standard forwarding-rule and load-balancer data-processing charges are not automatically covered by Compute Engine Committed Use Discounts (CUDs). These commitments target eligible compute resources rather than the load-balancing service itself.

However, eligible Compute Engine instances behind a load balancer may qualify for CUDs. This means you can optimize network costs separately while reducing the cost of supported backend infrastructure.

Learn more in our GCP Committed Use Discounts guide.

How We Help Optimize Eligible GCP Compute Costs

At Usage.ai, we help organizations manage eligible cloud discount commitments, including supported Google Cloud compute commitments.

For applications behind GCP Load Balancers, our focus is eligible backend compute spending rather than forwarding-rule or data-processing charges. We don’t replace Google’s load-balancer configuration, network monitoring, or traffic-management tools.

Our automated GCP CUD management guide explains how commitment optimization works.
GCP SAVINGS DEMO
Cut Your GCP Compute Costs

Find savings on eligible backend compute with automated commitment management. No infrastructure changes needed.

Final Verdict

GCP Load Balancer costs depend on forwarding rules, traffic processing, and, for internal Application Load Balancers, managed proxy capacity. The best way to control spending is to choose the appropriate configuration, eliminate unused resources, and optimize backend compute separately.

Frequently asked questions

How much does a GCP load balancer cost per month?

An external load balancer with one forwarding rule can start at approximately $18.25/month before traffic-processing charges. The final cost depends on traffic volume, architecture, and additional services.

Is Google Cloud Load Balancing free?

Generally, no. Forwarding rules and applicable processing or proxy resources generate charges, although promotional credits may offset eligible spending.

Does GCP charge for an idle load balancer?

Yes. Provisioned forwarding rules remain billable without traffic. Internal Application Load Balancers also retain minimum proxy charges.

How much does an internal Application Load Balancer cost?

In Iowa, the minimum three proxy instances cost $0.075/hour, or $54.75/month for 730 hours. Data processing and additional capacity can increase the bill.

Is a global load balancer more expensive than a regional one?

Not necessarily. Both can have similar forwarding-rule charges, but network service tiers, internet egress, and deployment requirements affect the overall cost.

Share
Facebook
X
LinkedIn
Reddit
See your savings first
More from Usage.ai
Latest from our blogs