Does Datadog have a free tier in 2026?
Yes. Datadog offers Infrastructure Free for up to five hosts, with one-day metric retention. The $0 plan is separate from the free trial. It should not be read as free access to all products, such as logs and APM. See the official infrastructure plan comparison.
For a small workload, start by deciding what you need to investigate: host health, individual requests, application logs, or browser behaviour. A plan that covers the first may not cover the others. Compare the required features and retention before deciding whether a free plan meets your needs.
We build Telemtra, an observability product. This guide links to Datadog's own pricing and billing documentation, checked on September 8, 2026. Quoted paid rates are USD starting prices with annual billing; monthly, on-demand, and negotiated terms can differ.
Seven areas to check in a Datadog estimate
These are common observability cost categories, not an exhaustive product list. Build your estimate from the products you intend to use, rather than multiplying one advertised host price by your server count.
1. Infrastructure monitoring and containers
Infrastructure Pro starts at $15 per host per month; Enterprise starts at $23. Containers have plan-specific allowances and can incur additional charges beyond them. Check the infrastructure pricing and container allowances.
For autoscaling, inspect the metering rule as well as the host count. Datadog's Usage Details documentation reports infrastructure hosts using the 99th percentile of hourly distinct-host counts. A brief peak is not automatically a permanent increase in next month's bill. Use your agreement's billing rules when modelling a bursty workload.
2. APM and trace volume
The public price list distinguishes APM Pro ($35 per APM host per month) from APM Enterprise ($40). Do not label the base APM tier as Pro.
Host fees are only part of the estimate. Check included ingested and indexed spans, retention, and overages in the product allotments. Review which services require tracing and how sampling affects the investigations you need to perform. Cutting collection indiscriminately can remove the evidence needed to diagnose an incident.
3. Log ingestion, indexing, and retention
Datadog separates log ingestion from indexing. Its Logging without Limits guide explains how to collect logs while choosing which ones need indexing. Estimate incoming bytes and indexed event counts separately, using the retention period you actually need.
An index exclusion filter does not undo ingestion. It prevents matching logs from being indexed. To reduce the volume sent to Datadog, filter or reduce logs before export. Review index filters, exclusion filters, and quotas when planning searchable retention.
Start with repetitive health checks and debug output. Confirm that any proposed exclusion preserves the errors, audit records, and diagnostic context your team needs.
4. Custom metric series
Custom metric usage depends on metric names and combinations of tag values, including the host tag. Datadog measures usage hourly and uses a monthly average; indexed and ingested custom metrics can have different billing treatment. See custom metrics billing for the applicable model.
Review unbounded tags such as user_id, request_id, or raw URL paths. Before adding one, estimate how many distinct values it could create and whether that detail belongs in a log or trace instead. There is no single savings percentage that applies to every application.
5. Real User Monitoring and Session Replay
Current pricing distinguishes RUM Measure, RUM Investigate, and Session Replay. These have different session-based meters; the old “RUM Lite” shorthand is not a reliable estimate. Consult the RUM price-list entries.
Check collection and replay settings against the traffic you expect. Datadog documents browser session sampling, including sessionSampleRate. Compare the resulting usage with your plan rather than assuming every crawler is a billable browser session.
6. Synthetic monitoring
Listed annual rates start at $5 per 10,000 API test runs and $12 per 1,000 browser test runs. Check the synthetics pricing entries and your agreement.
For example, one browser test scheduled every five minutes from ten locations produces 2,880 scheduled runs per day. Over 30 days, that is 86,400 runs. At $12 per 1,000, the arithmetic is $1,036.80 per month, before retries, discounts, or other charges. This is an illustrative usage estimate, not a quote.
Choose test frequency and locations to match the failures you need to detect. Recalculate the estimate whenever you add a location or shorten the interval.
7. Continuous Profiler
Continuous Profiler is included in APM Enterprise; the standalone listed rate is $19 per profiled host per month with annual billing. Avoid counting both a bundle and its included feature as separate charges. See the APM and profiler entries.
Review the APM billing documentation for the host or task model that applies to your deployment. Decide where profiling is useful before estimating the number of profiled workloads.
A practical cost audit checklist
- List the enabled products, billing units, commitments, and overage rates in your agreement.
- Compare actual usage with included allowances for each product.
- Identify metric tags that create large numbers of series.
- Separate log ingestion volume from indexed volume and retention.
- Check tracing, profiling, RUM, and replay collection settings.
- Multiply synthetic test frequency by locations and days in the billing period.
- Make one collection change at a time, then check both usage and diagnostic coverage.
Use Datadog's estimated usage metrics to watch for changes during the month. Record the deployment date of each adjustment so you can distinguish its effect from a traffic spike or a new service.
Comparing Datadog with alternatives
Compare the same workload, retention, and required features across vendors. Include allowances, overages, support, and the engineering effort to operate the setup. A lower advertised starting price does not establish a lower total cost.
For a Telemtra comparison, use our current pricing and test with a representative workload. Create a Telemtra workspace to evaluate whether it fits the investigations your team needs to perform.
Frequently asked questions
Does Datadog have a free tier in 2026?
Yes. Infrastructure Free costs $0 and supports up to five hosts with one-day metric retention. This limited plan is different from the free trial; it is not a free subscription to every Datadog product.
What makes Datadog bills grow unexpectedly?
More monitored hosts, additional custom metric series, higher log volumes, and more synthetic test runs can increase usage. Check each product's billing unit, retention, included allowance, and overage rate before changing collection settings.
Do log exclusion filters eliminate ingestion charges?
No. Index exclusion filters control which ingested logs are indexed. Reducing ingestion requires reducing the data sent to Datadog, for example by filtering it at the source or in a collector before export.
Further reading
See Datadog alternatives in 2026, how to avoid noisy observability bills, and our Data Shaper documentation for more context on evaluating and controlling telemetry volume.