Skip to content

Billing FAQ

How is usage calculated?

Different data types are calculated differently each billing period:

Data type Method
Metrics Mean average over the billing period
Logs Sum of all data ingested
Traces Sum of all data ingested
OpsPilot tokens Sum of all tokens used
FR hours Sum of all instance hours

Why is my bill higher than expected?

The most common causes are:

  • Metrics cardinality — A spike in unique metric series (e.g. a deployment creating many new label combinations) can push your mean average up.
  • Log volume — Verbose log levels (DEBUG, TRACE) generate significantly more data than INFO or WARN.
  • Trace volume — A high-traffic application with 100% sampling will ingest traces rapidly.
  • AI token usage — Long, multi-turn OpsPilot AI conversations consume tokens quickly as previous context is included in each request.

Use the Billing usage dashboard for a full breakdown of where costs are coming from.

How do I reduce costs?

FusionReactor agent

Reduce the trace sampling ratio using the JVM property:

-Dfr.observability.trace.sampling.ratio=0.03

Common values:

Ratio Sampling rate Recommended for
1.0 100% Development / critical services
0.1 10% Standard production
0.05 5% Default
0.03 3% High-volume applications
0.01 1% Very high-volume, cost-sensitive

Grafana Alloy / OTel collector

Optimize at the collector level:

  • Traces — Configure a tail sampling processor to drop low-value spans.
  • Logs — Add a filter stage to drop DEBUG/TRACE log levels or noisy sources before they are sent.
  • Metrics — Increase scrape intervals or use relabelling to drop unused metric series.

OpsPilot AI tokens

  • Keep queries focused and concise — long conversations accumulate context quickly.
  • Start a new conversation rather than extending a very long one.

What happens when I exceed my plan limits?

Usage beyond your plan's included allowance is billed at on-demand rates. There is no hard cutoff — data continues to be ingested and you are charged for the overage at the end of the billing period.

How do I set up a billing alert?

You can create an alert using the data usage metrics to notify you before you exceed your allowance:

  • fr_logs_bytes_received — alert when log ingestion approaches your plan limit
  • fr_traces_bytes_received — alert when trace ingestion approaches your plan limit
  • fr_metrics_series_count — alert when your metric series count approaches your plan limit
  • fr_usage_minutes — alert when FR instance hours approach your seat allowance

Learn more

Create an alert rule

What is the difference between annual and monthly pricing?

The annual plan is approximately 20% less expensive than monthly. Both are billed monthly, but the annual plan requires a 12-month commitment.

What counts as a seat?

A seat covers a single unique virtual or physical host with up to 4 instances of Java/ColdFusion installed, or up to 4 Docker containers.

Example

5–8 instances on a host requires 2 seats.

How do I upgrade or downgrade my plan?

Upgrades take effect immediately. Downgrades take effect at the end of the current billing period.

Manage your plan from Administration > Billing in the OpsPilot Portal.

When am I charged?

Both annual and monthly plan usage is calculated and billed each month.

How do I cancel?

Contact support via the chat bubble in the OpsPilot Portal.

Need more help?

Contact support in the chat bubble and let us know how we can assist.