You are currently viewing AWS Cost Explorer: A Practical Guide to Reports, Forecasts and Limits

AWS Cost Explorer: A Practical Guide to Reports, Forecasts and Limits

Spread the love

Updated November 2026.

AWS Cost Explorer is the free reporting tool inside AWS Billing and Cost Management. It charts cost and usage over time, lets you filter and group by service, account, Region, tag and usage type, forecasts future spend, and recommends Savings Plans and Reserved Instance purchases. For most teams it is the first place to answer “why did the bill go up?”

It changed more in the last year than in the five before. Forecasts now reach 18 months, and an Amazon Q button explains cost changes in plain language. It still has hard limits, though, and knowing them stops you from arguing with a chart that was never built for your question.

Turning it on

Open the Billing and Cost Management console and choose Cost Explorer. You cannot enable it through the API. Once enabled, AWS prepares your cost data for the current month and the last 12 months and calculates a forecast; current month data shows up in about 24 hours.

In AWS Organizations, the management account sees every member account. Member accounts see their own spend unless the management account restricts it.

What Cost Explorer can show you

CapabilityWhat you get
HistoryUp to 38 months at monthly granularity, for year-over-year analysis
GranularityDaily and monthly by default; hourly for the past 14 days when enabled
Resource-level dataDaily for all services, hourly for EC2, when enabled
ForecastUp to 18 months ahead, using up to 36 months of history
AI explanations“Analyze with Amazon Q” for historical, forecast and comparison views
RecommendationsSavings Plans and Reserved Instance purchase suggestions
ReportsCoverage and utilization for commitments

Hourly and resource-level data are optional settings and carry a charge, so turn them on for the accounts where you need them rather than everywhere.

The reports worth knowing

Monthly costs by service

The default view. The top three or four services usually account for most of the bill, so start there.

Group by usage type

When a service looks too large, group it by usage type. EC2 splits into compute, EBS, data transfer, NAT gateway and public IPv4 charges; S3 splits into storage classes, requests and transfer. Most surprises live here.

Group by linked account, then by tag

Linked accounts show which team or environment owns a spike. Tags add detail once you activate them as cost allocation tags, but only from that moment onward; past spend stays unattributed. Expect up to a day before tagged data appears. Report the untagged share every month. Cloud cost allocation explains why.

Group by purchase option

Shows On-Demand, Savings Plans, Reserved and Spot spend side by side, which is the quickest way to see how much eligible usage is still uncommitted. Reserved vs Spot vs On-Demand covers what to do with that number.

AWS Cost Explorer chart grouped by service with daily and monthly views

Daily view for spikes

Switch to daily granularity to tell a steady climb from a one-day spike. A step change that stays is usually a deployment; a spike that returns to baseline is usually a job or a test.

Forecasts: better, still not a budget

In November 2025, AWS extended Cost Explorer forecasts from 12 to 18 months and upgraded the models to analyze up to 36 months of history, up from six, to pick up seasonality and long-term growth. The 18-month horizon is available in the console and through the GetCostForecast API.

AWS also added AI-powered explanations of the drivers behind each forecast, initially in preview and only in the console.

Treat the forecast as a trend line. It sees your history, not your roadmap: a product launch, a migration or a new AI feature will not appear until it shows up in usage. FP&A teams should combine it with planned changes; see the FP&A process.

Analyze with Amazon Q

Since June 9, 2026, Cost Explorer has an “Analyze with Amazon Q” button that explains whatever report you are viewing. AWS says it adapts to the date range: historical explanations for past periods, forecast explanations for future ones, or both for mixed ranges. AWS’s walkthrough shows it drawing on pricing data, budgets, anomaly detection and optimization recommendations. AWS lists it at no additional cost in all commercial Regions.

It is most useful for people who do not live in billing data: an engineering manager asking why their account rose 18% in May gets a written answer with the top drivers and their share of the change. Check the numbers it cites against the chart before you forward them; it explains, it does not audit.

Commitment recommendations

Cost Explorer recommends Savings Plans and Reserved Instances based on past usage, with projected savings. That is also its weakness: a workload you plan to retire next month still looks worth committing to. Before buying:

  1. Check the lookback period the recommendation uses, and compare it with planned changes.
  2. Recalculate after cleanup and rightsizing; see the cloud cost optimization checklist.
  3. Read Reserved Instances versus Savings Plans.
  4. After buying, watch utilization and coverage reports monthly.

What it costs

The console is free. The Cost Explorer API charges per request ($0.01 per paginated request on the AWS pricing page at the time of writing), which is trivial for a weekly report and noticeable for a dashboard that polls constantly. Hourly and resource-level granularity are billed separately. Check the pricing page for current rates.

NeedConsoleAPI
Ad hoc investigationBest fitNot ideal
Scheduled reportsSaved reports, CSV exportScripts or dashboards
Feed a BI or FinOps toolNot possiblePossible, per-request cost
Detailed line itemsNot possibleUse Data Exports instead

Where it stops being enough

Not real time. Cost data lags by hours. To catch a runaway job, use AWS Budgets alerts and Cost Anomaly Detection.

No split container costs. AWS split cost allocation data for EKS and ECS, which divides instance cost by pod or task, appears in the Cost and Usage Report, not in Cost Explorer. Kubernetes cost monitoring covers the alternatives.

Limited shared cost logic. Cost Categories and split charge rules help, but complex allocation and per-customer unit costs push you toward Data Exports (CUR 2.0 or FOCUS 1.2) queried with Athena or a FinOps platform.

One cloud only. For AWS plus Azure or Google Cloud, see multi-cloud cost management.

Cost Explorer limits and where to use Budgets, anomaly detection and the Cost and Usage Report

A first-hour routine

  1. Open Monthly costs by service and note the top four services.
  2. Group the biggest by usage type and look for data transfer, NAT, IPv4 and log charges.
  3. Group by linked account, then by tag, and write down the share with no owner.
  4. Group by purchase option to see uncommitted eligible usage.
  5. Click “Analyze with Amazon Q” on last month’s report and compare its drivers with your own reading.
  6. Check commitment coverage and utilization.
  7. Set a budget and anomaly alert so the next spike reaches you before the invoice.

Then work through the findings with how to reduce your AWS bill, and estimate changes with the AWS cost calculator.

Questions people ask

Is AWS Cost Explorer free?

The console is free. API requests carry a small per-request charge, and optional hourly and resource-level data are billed.

How far ahead does Cost Explorer forecast?

Up to 18 months, since November 2025, using up to 36 months of history.

Why do my tags not show up?

Tags must be activated as cost allocation tags, and only spend after activation carries them. Allow up to a day.

Can Cost Explorer show Kubernetes costs per namespace?

Not directly. Split cost allocation data for EKS is available in the Cost and Usage Report, not in Cost Explorer.

Sources

About the Author

Nhon Dang is a cloud infrastructure and operations professional with over 10 years of hands‑on experience across AWS services, Kubernetes, object storage, managed databases, and GPU infrastructure. He writes practical, experience‑driven guidance to help teams build reliable, scalable, and cost‑efficient cloud environments.

Nhon Dang

Nhon Dang is a cloud infrastructure and operations professional with over 10 years of hands-on experience in cloud services, infrastructure, and business operations. His expertise spans the design, deployment, and operation of cloud platforms and managed services, including virtual machines (VMs), Kubernetes (K8s), object storage (S3), managed databases, Apache Kafka, and cloud GPU infrastructure. Throughout his career, Nhon has worked closely with cloud infrastructure and service operations, gaining practical experience in building reliable, scalable, and cost-efficient cloud environments. His work combines technical expertise with business and operational insight, giving him a practical perspective on how cloud technologies perform in real-world production environments. Nhon writes about cloud infrastructure, Kubernetes, DevOps, distributed systems, cloud computing, infrastructure operations, and cloud service management, sharing insights based on hands-on experience rather than purely theoretical knowledge. His goal is to provide practical, technically accurate, and experience-driven guidance that helps engineers, technical teams, and businesses make better decisions when adopting and operating cloud technologies.

This Post Has 3 Comments

Leave a Reply