AWS CodeBuild Pricing Calculator
Estimate CodeBuild compute minutes, ARM savings, and cache storage before your pipeline bill surprises finance.
- No signup
- Instant results
- ~2 min
Rates reviewed
Read the full breakdown — Companion pricing guide covers billing dimensions, traps, and a 30-day cleanup plan.
Compute fleet
us-east-1list rates, pricing as of 2026-06-30. CodePipeline, NAT egress, CloudWatch Logs, and artifact S3 excluded. Source: AWS public pricing.
Who This Tool Is For
DevOps and platform engineers forecasting CI/CD spend as monorepo build frequency and Docker image sizes grow.
Why We Built This Tool
CodeBuild bills per build minute (or per second on Lambda compute) — a misconfigured LARGE environment on every lint step compounds faster than commit velocity.
What Problem It Solves
- Oversized compute. BUILD_GENERAL1_LARGE on lint jobs that need SMALL wastes 4× per minute.
- Missing cache. Re-downloading node_modules on every build burns minutes that S3 cache eliminates.
- NAT blind spot. VPC builds pulling public dependencies add NAT Gateway charges not shown here.
Read cost-aware CI/CD or explore DevOps pipeline setup.
Frequently Asked Questions
Where do these CodeBuild prices come from?
AWS published us-east-1 list prices (AWS CodeBuild pricing), pricing as of 2026-06-30.
What is the CodeBuild free tier?
100 build minutes per month on general1.small or arm1.small. Lambda compute gets 6,000 build seconds per month on lambda.arm.1GB or lambda.x86-64.1GB.
How much does Graviton save?
arm1.small is $0.004/min vs $0.005/min on general1.small — 20% cheaper. arm1.large is $0.016/min vs $0.02/min on general1.large.
When should I use Lambda compute for builds?
Lambda compute suits lightweight lint, test, and script steps without Docker — roughly 10× cheaper per second than general compute for short jobs.
Does CodeBuild include CodePipeline charges?
No. CodePipeline V2 bills $0.002 per action execution minute separately. This calculator covers CodeBuild compute and S3 cache storage only.
How do I reduce build minutes?
Enable S3 or local cache for dependencies, right-size compute per step, parallelize independent stages, and use Lambda compute for non-Docker steps.