Datadog’s pricing is not one meter. Its official pricing pages and billing documentation show separate charging models across infrastructure monitoring, logs, application performance monitoring, security, developer tooling, and incident response. For B2B buyers, that matters because one environment can generate charges by host, by gigabyte, by event, by session, by committer, and by seat in the same month.
What it is and who it’s for
Datadog sells a broad observability and operations platform rather than a single product. Its pricing page lists product-specific models across infrastructure, logs, security, APM, and developer tools.
That makes the platform relevant to several buying centers at once. Infrastructure teams may start with Infrastructure Pro or Enterprise, application teams may add APM, platform teams may turn on Container Monitoring or Cloud Network Monitoring, security teams may buy Cloud SIEM or Cloud Security Management, and developer productivity teams may use Pipeline Visibility, Test Optimization, or Code Coverage.
Datadog also prices some products around consumption by business workflow rather than only monitored machines. Examples on the comparison page include On-Call per seat, Incident Management per seat, Product Analytics per 1,000 sessions, Synthetic API Tests per 10,000 test runs, and Workflow Automation per 100 workflow executions.
How it works
According to Datadog’s billing documentation, fees are calculated based on product usage during each calendar month unless an order says otherwise. That monthly usage is then measured differently depending on the product.
For host-based products, Datadog offers two billing plans: a High watermark plan and a Hybrid monthly/hourly plan. Datadog says the High watermark plan bills on the maximum count of the lower 99% of hourly readings, excluding the top 1% of hourly usage to reduce the impact of spikes. Datadog says the Hybrid monthly/hourly plan combines a monthly minimum with hourly charges above commitment.
The metering unit also changes by product family. Infrastructure hosts are counted hourly as unique monitored hosts, while containers are recorded every five minutes and billed on fractional hours. AWS Fargate tasks are also measured at five-minute intervals and billed by total monitored hours. APM host billing follows the same high-watermark and hybrid mechanics as Infrastructure.
Usage-priced products add another layer. Datadog says Log Management charges ingested logs by total gigabytes submitted, while indexed log events are billed per million logs at the rate for the selected retention policy. For APM, ingested spans are billed by total gigabytes of spans ingested into APM, and indexed spans are billed by total spans indexed by retention filters. For RUM, or Real User Monitoring, sessions are the billable unit, and a session expires after 15 minutes of inactivity or 4 hours of continuous activity.
Some products use definitions that buyers should verify early. Datadog says a CI Visibility committer counts only if that person commits at least three times in a month, and it counts unique committers who send test and pipeline data. In Incident Management legacy usage-based plans, a user counts as active if they contribute substantively to incident response.
Pricing and cost considerations
The fastest way to understand Datadog pricing is to separate base monitoring charges from usage and retention charges that may sit on top of them.
| Product or feature | Billing unit | Published price / billing rate |
|---|---|---|
| Infrastructure Pro | Per infra host per month | $15 |
| Infrastructure Enterprise | Per infra host per month | $23 |
| Infrastructure DevSecOps Pro | Per infra host per month | $22 |
| Infrastructure DevSecOps Enterprise | Per infra host per month | $34 |
| Container Monitoring | Per container per month or per container hour | $1 per container per month; $0.002 per container hour |
| Cloud Network Monitoring | Per network host per month | $5 |
| Network Device Monitoring | Per network device per month | $7 |
| APM | Per APM host per month | $31 |
| APM Pro | Per APM host per month | $35 |
| APM Enterprise | Per APM host per month | $40 |
| APM Ingestion | Per ingested GB per month | $0.10 |
| APM Indexed Spans — 7-day retention | Per 1M indexed spans per month | $1.27 |
| APM Indexed Spans — 15-day retention | Per 1M indexed spans per month | $1.70 |
| APM Indexed Spans — 30-day retention | Per 1M indexed spans per month | $2.50 |
| Universal Service Monitoring | Per infra host per month | $9 |
| Data Streams Monitoring | Per host per month | $15 |
| Continuous Profiler | Per profiled host per month | $19 |
| Database Monitoring | Per database host per month | $70 |
| Logs ingestion | Per ingested logs (1GB) per month | $0.10 |
| Standard Indexing | Per million log events per month with 15-day retention | $1.70 |
| Flex Storage | Per 1M events stored per month | $0.05 |
| Flex Logs Starter | Per 1M events stored per month | $0.60 |
| Log Forwarding to Custom Destinations | Per GB outbound, per destination, per month | $0.25 |
| Cloud SIEM | Per 1M analyzed events per month | $5 |
| Managed Platform OTLP | Per GB of ingested spans per month | $0.50 |
| Custom Metrics | Per 100 custom metrics per month | $5 |
| Ingested Custom Metrics | Per 100 ingested custom metrics per month | $0.10 |
| Custom Events | Per 100K custom events per month | $2 |
| RUM Measure | Per 1,000 sessions per month on full traffic | $0.15 |
| RUM Investigate | Per 1,000 sessions per month on filtered sessions | $3 |
| Session Replay | Per 1,000 sessions per month | $2.50 |
| Product Analytics | Per 1,000 sessions per month on full traffic | $0.80 |
| Synthetic API Tests | Per 10K API test runs per month | $5 |
| Synthetic Browser Tests | Per 1K browser test runs per month | $12 |
| Mobile App Testing | Per 100 test runs per month | $50 |
| Pipeline Visibility | Per active Git committer per month | $8 |
| Test Optimization | Per active Git committer per month | $20 |
| Code Coverage | Per active Git committer per month | $8 |
| On-Call | Per seat per month | $20 |
| Incident Management | Per seat per month | $30 |
| Incident Response | Per seat per month | $40 |
| Workflow Automation | Per 100 workflow executions per month | $10 |
| AI Credits | Per 500-credit bundle per month, or per credit on-demand | $500 per 500 credits per month; $1.30 per credit on-demand |
The figures above are published starting rates or usage rates; on-demand prices can be higher for some products. APM host rates do not represent the full APM bill — ingestion and indexed spans are separate usage-based charges. Why the bill compounds becomes clearer when those units are viewed together. A single production service might sit on infrastructure hosts, emit logs that are charged once for ingestion and again for indexing or storage, send traces into APM, and also produce RUM sessions from the user side. Datadog’s documentation also makes clear that indexed and ingested span volumes are separate billable dimensions in APM, just as ingestion and indexed events are separate dimensions in Log Management.
Retention is another multiplier. Datadog’s pricing pages say indexed log events have tiered retention pricing from 3-day to 30-day retention, APM indexed spans have tiered retention pricing from 7-day to 30-day retention, and NetFlow Monitoring pricing increases with retention length. In logs, Datadog says on-demand Log Management defaults to 15 days retention, while Flex Logs Starter includes 3-month, 6-month, 12-month, and 15-month retention options.
Commitment structure also affects the effective bill. Datadog says pricing is shown as “starting at” and may vary by commitment or usage.
There are also explicit overage mechanics. Datadog says unexpected log spikes are not rate-limited by a single day’s volume, and extra services may be billed at 50% above the annual price for volume above commitment.
How to choose
Buyers usually get the clearest answer by mapping Datadog products to the team behavior that creates the bill.
If the main need is host and container visibility, the first question is whether billing will track infrastructure hosts, containers, or both. Datadog counts hosts hourly and containers every five minutes, so environments with rapid scaling patterns should model both unit types before committing to Infrastructure Monitoring and Container Monitoring together.
If logs are central, separate three decisions that Datadog prices independently: ingestion volume in gigabytes, indexed event volume, and longer-term storage or forwarding. Datadog also says Log Management can be bought without Infrastructure or APM, which matters for teams that only want centralized logs.
If application tracing is the driver, APM buyers should distinguish host charges from ingestion and indexed span charges. Datadog’s documentation says controls can be put in place for both indexed and ingested span volumes, so procurement and platform teams should ask how those controls will be configured before turning tracing on broadly.
For developer productivity products, the important question is who counts as a billable contributor. Datadog defines a CI Visibility committer as someone who commits at least three times in a month, and Pipeline Visibility counts every pipeline, stage, and job as a pipeline span.
Limitations and gotchas
Datadog’s biggest pricing challenge is not that it has one complicated tariff. It is that the platform has many separate tariffs that can apply to the same workload.
The first gotcha is partial budgeting. A team may price only infrastructure hosts and miss add-on charges for APM ingestion, indexed spans, logs ingestion, indexed log events, or session-based products such as RUM and Session Replay.
The second is that “host-based” does not always mean a simple fixed monthly count. Datadog’s own billing docs say host products can use high-watermark or hybrid monthly/hourly billing, and the exact hourly pricing above commitment is not specified on the billing page.
The third is feature trade-offs across storage options. Datadog says Flex Logs does not support monitors or Watchdog Insights.
The fourth is outbound duplication cost in logging architectures. Datadog lists Log Forwarding to Custom Destinations as billed per outbound GB, per destination, per month, so sending the same data to multiple destinations creates multiple billable paths.
The fifth is hidden complexity in “unlimited” language. Datadog says Log Management includes unlimited sources and unlimited user accounts, but those allowances do not remove ingestion, indexing, storage, and forwarding charges inside the same logging setup.
The sixth is regional availability constraints. Datadog notes that several products, including Feature Flags and Product Analytics, are not currently supported in GovCloud environments, and Disaster Recovery is also unavailable in GovCloud (us1.fed).
FAQ
Do unused AI Credits roll over to the next month?
No. Datadog says AI Credits reset on the first day of each calendar month, and unused credits do not roll over to the next month.
Can I forward logs from one Datadog account to another?
No. Datadog states that forwarding logs from one Datadog account to another is currently not supported.
How are subtests billed in Synthetic Browser and Mobile App Testing?
Datadog treats subtests as part of the overall test execution. For example, three subtests that contain five steps each are billed as a total of 15 steps.
What retention does Datadog Log Management use for on-demand customers?
Datadog says the default retention for on-demand customers is 15 days.
Does Flex Logs support monitors and Watchdog Insights?
No. Datadog says Flex Logs does not support monitors or Watchdog Insights, while Standard Indexing supports these features.
