Pricing 10 min read

Datadog pricing, explained

What a host, a gigabyte of logs and a custom metric cost, and three monthly bills worked out line by line.

JeremyFunk By JeremyFunk Prices verified 2026-11-21

Datadog is a monitoring platform sold as separate products: infrastructure, APM, log management, real user monitoring, synthetics and dozens more. Each product has its own unit, so a bill adds up hosts, gigabytes, million-event blocks, sessions and test runs.

This guide covers how the products are billed, the list prices for the ones most teams use, the charges that surprise people, three worked monthly bills, and how to bring a Datadog bill down.

Maple publishes this guide and competes with Datadog. Section 06 compares the two; every Datadog figure comes from Datadog's own pages, linked at the end.

Key takeaways

  • Infrastructure Pro is $15 per host per month on an annual plan, $18 on-demand.
  • APM is another $31 per APM host on an annual plan, $36 on-demand.
  • Logs bill twice: $0.10 per GB ingested, then $1.70 per million events indexed for 15 days.
  • Each tag combination on a metric is a separate custom metric. Pro includes 100 per host.
  • Hosts are counted hourly, and the bill uses the highest count after dropping the top 1% of hours.

01 · The model

How Datadog pricing works

Datadog has no single plan. You turn on products, and each one meters its own unit. The ones in this guide bill in five ways:

CategoryUnitExamples
HostsPer host per monthInfrastructure, Database Monitoring
APM and spansPer APM host, plus spans per GB and per millionAPM, indexed spans
LogsPer GB ingested, plus per million events indexedLog Management, Flex Logs
Custom metricsPer 100 metrics above the host allotmentMetrics from your code or OpenTelemetry
RUM and syntheticsPer 1,000 sessions, per test runRUM, Session Replay, API and browser tests

The free tier

Datadog's free tier covers up to five hosts with one day of metric retention. It is enough to try the agent and the dashboards. APM, logs and the other paid products are not part of it.

Annual, monthly and on-demand

Every product has three prices: annual, month-to-month and on-demand. Annual is the lowest and needs a commitment to a volume for the year. On-demand needs no commitment and costs more. For hosts the gap is 20%: Infrastructure Pro is $15 annual and $18 on-demand. For 15-day log indexing it is 50%: $1.70 against $2.55 per million events. For logs, the pricing page says a month above your committed volume pays 50% over the annual price, only on the volume past the commitment. A commitment sized below real usage pays that premium on the rest.

How hosts are counted

Datadog's billing docs say hosts are metered every hour. At the end of the month it drops the top 1% of hours and bills the highest remaining count, the high-water mark of the lower 99%. A fleet that runs 20 hosts most of the month and scales to 60 for two days pays for 60, because two days is more than 1% of a month. A container running its own agent counts as a host; the docs recommend one agent per host for that reason.

Verified 2026-11-21 against datadoghq.com/pricing, the full price list and Datadog's billing docs.

02 · List prices

The Datadog price list

These are the US list prices for the products most teams start with. The full list has more than a hundred lines.

ProductUnitAnnualOn-demand
Infrastructure ProPer host per month$15$18
Infrastructure EnterprisePer host per month$23$27
APMPer APM host per month$31$36
APM EnterprisePer APM host, adds the profiler$40$48
Log ingestPer GB$0.10$0.10
Log indexing, 15 daysPer million events$1.70$2.55
Custom metricsPer 100, above the allotment$5$5
RUM MeasurePer 1,000 sessions$0.15$0.22
RUM InvestigatePer 1,000 retained sessions$3$4.50
Session ReplayPer 1,000 sessions$2.50$3.60
Synthetic API testsPer 10,000 runs$5$7.20
Synthetic browser testsPer 1,000 runs$12$18
Database MonitoringPer database host per month$70$84
Error TrackingFlat price for the first 50,000 errors a month$25$36

Month-to-month prices sit between the two columns for most products and match on-demand for hosts. Existing customers may have different allotments in their contract.

Each Infrastructure Pro host includes 100 custom metrics and 5 containers; Enterprise includes 200 and 10. Extra containers cost $1 a month each prepaid or $0.002 per container-hour on-demand. Allotments are pooled across the account, so one busy host can use the allotment of a quiet one.

Logs: ingest, then index

Every log Datadog receives pays $0.10 per GB to ingest. Logs you want to search and alert on go into an index, which bills per million events at the retention you choose:

Index retentionAnnual, per million eventsOn-demand
3 days$1.06$1.59
7 days$1.27$1.91
15 days$1.70$2.55
30 days$2.50$3.75

Indexing is priced per event, not per byte. At 1 KB per log line, a GB holds about a million events, so indexing it for 15 days costs 17 times what ingesting it did. Flex Logs is a cheaper storage tier for logs you search rarely, at $0.05 per million events stored per month on an annual plan.

APM and spans

Each APM host includes 150 GB of ingested spans and 1 million indexed spans a month, both pooled across the account. Ingested spans above the allotment are $0.10 per GB. Indexed spans, the ones kept for search and analytics, are $1.70 per million for 15 days on an annual plan or $2.55 on-demand. On on-demand billing, the pricing page says span allotments are metered each hour rather than pooled over the month, so a traffic spike can cross the allotment even when the monthly total would not.

03 · Surprises

What surprises teams on a Datadog bill

  • Custom metrics multiply with tags. Datadog counts every distinct combination of metric name and tag values, including host, as one custom metric. A distribution metric counts five per combination, and turning on percentiles doubles that. The worked example below prices one such metric.
  • Indexed logs cost more than ingested ones. Teams budget for $0.10 per GB and then index most of what they send.
  • Short spikes set the host count. Anything above 1% of the month, about 7 hours, counts toward the high-water mark. Autoscaling for a weekly batch job can set the bill.
  • Containers past the allotment. A Kubernetes node packed with small pods passes 5 containers per host quickly, and each extra one bills by the hour.
  • Each product is its own line. RUM, Session Replay, synthetics, Database Monitoring, error tracking and security products each add a meter. A dashboard that pulls from five products is billed by five.
  • Growth past the commitment costs more per unit. On-demand prices run about 16% to 50% above annual ones, and log volume above a commitment pays 50% over the annual price.

04 · Examples

What a month of Datadog costs

Three bills at list price, from rates checked 2026-11-21. Volumes are monthly and a month is 30 days. Each bar splits one bill by category:

Small team: about $208.40 a month

Five hosts on Infrastructure Pro, three of them running APM. They send 50 GB of logs, about 50 million events, and index 10 million of them for 15 days. Their services emit 800 custom metrics and they index 5 million spans. Annual prices:

LineCalculationPer month
Infrastructure Pro5 hosts × $15.00$75.00
APM3 APM hosts × $31.00$93.00
Log ingest50 GB × $0.10$5.00
Log indexing, 15 days10M events × $1.70$17.00
Custom metrics800 − 500 included = 300, × $5.00 per 100$15.00
Indexed spans5M − 3M included = 2M × $1.70$3.40
Total$208.40

Hosts and APM are 81% of this bill. The 50 GB of logs cost $22.00 in total, less than APM for one host.

Mid-sized team: about $3,056.50 a month

Forty hosts on Infrastructure Pro, 25 of them running APM. They ingest 1,000 GB of logs and index 200 million events for 15 days. Services send 600 GB of spans, emit 12,000 custom metrics and index 60 million spans. The web app has 500,000 sessions a month on RUM Measure, with Session Replay on 50,000 of them, which are retained on RUM Investigate so they can be replayed. Ten API checks run every minute from two locations, 864,000 runs.

LineCalculation, annualAnnualOn-demand
Infrastructure Pro40 hosts × $15.00$600.00$720.00
APM25 APM hosts × $31.00$775.00$900.00
Log ingest1,000 GB × $0.10$100.00$100.00
Log indexing, 15 days200M events × $1.70$340.00$510.00
Custom metrics12,000 − 4,000 included = 8,000, × $5.00 per 100$400.00$400.00
Indexed spans60M − 25M included = 35M × $1.70$59.50$90.12
RUM Measure500k sessions × $0.15 per 1,000$75.00$110.00
RUM Investigate50k replayed sessions retained × $3.00 per 1,000$150.00$225.00
Session Replay50k sessions × $2.50 per 1,000$125.00$180.00
Synthetic API tests864,000 runs × $5.00 per 10,000$432.00$622.08
Total$3,056.50$3,857.20

The same month on-demand costs $800.70 more. The on-demand span line assumes traffic is even across hours; since on-demand span allotments are metered hourly, uneven traffic costs more. The 864,000 synthetic runs cost almost as much as ingesting and indexing all the logs, because every location multiplies every check. The 600 GB of spans stays inside the 3,750 GB that 25 APM hosts include, so span ingest adds nothing here.

One custom metric: $2,000.00 a month

The mid-sized team adds one distribution metric, checkout.latency, tagged with endpoint (20 values), status (5) and host (40). If every combination reports in every hour of the month, the worst case:

StepCalculationCustom metrics
Tag combinations20 × 5 × 404,000
Distribution aggregations4,000 × 5 (count, sum, min, max, avg)20,000
With percentiles turned on20,000 × 240,000
Monthly cost40,000 ÷ 100 × $5$2,000.00

The team's 4,000 included metrics are already used, so all of it is billed. Dropping the host tag leaves 100 combinations, 1,000 custom metrics and $50.00 a month. This is the arithmetic from Datadog's own custom metrics billing docs, which use the same five-per-combination rule.

05 · Cutting the bill

How to cut a Datadog bill

  1. 01

    Find the top line first

    The Plan & Usage page breaks the bill down by product. In the mid-sized example, hosts and APM are about half; in a log-heavy setup, indexing is. Work on the biggest line.

  2. 02

    Index less, or index for less time

    Exclusion filters on an index stop events from being indexed; they still pay $0.10 per GB to ingest. Moving debug and access logs from a 15-day index to a 3-day one cuts their indexing price from $1.70 to $1.06 per million. Flex Logs suits logs you only search during an audit.

  3. 03

    Drop high-cardinality tags from metrics

    The Metrics Summary page lists custom metrics by count. Remove host, user IDs and request IDs from metric tags, and turn off percentiles you never chart. Metrics without Limits lets you choose which tags stay queryable.

  4. 04

    Keep APM on the hosts that need it

    APM is priced per host, so a utility host running the tracer costs $31 a month for traces nobody reads. Retention filters decide which ingested spans get indexed.

  5. 05

    Run one agent per node

    An agent inside each container makes every container a billable host. One containerized agent per Kubernetes node counts the node once and its containers against the allotment.

  6. 06

    Commit to the steady baseline

    Commit annually to the volume you always use and leave peaks on-demand. Size the commitment from a few months of the usage page, not from a host count at signing.

Synthetics can be a quick saving. A check that runs every 5 minutes instead of every minute is 80% fewer runs, at the cost of noticing an outage up to 4 minutes later.

06 · Maple

Maple pricing for the same workload

Maple is an OpenTelemetry observability platform. It has one plan at $39/month with 100 GB each of logs, traces and metrics included, then $0.30 per GB for each signal. There is no host, seat or index charge. Browser sessions with replay are 5,000 a month included, then $0.002 each. Logs and traces are kept 30 days and metrics 90.

The mid-sized team's month on Maple. Maple meters uncompressed OTLP bytes. Metrics are estimated at 0.1 KB per data point, one point a minute, for the 12,000 custom metrics plus about 200 system series per host (8,000 in all), which Datadog's host price covers:

LineCalculationPer month
PlanMonthly price$39.00
Logs1,000 GB − 100 included = 900 GB × $0.30$270.00
Traces600 GB − 100 included = 500 GB × $0.30$150.00
Metricsabout 86.4 GB, inside the 100 GB included$0.00
Session replays50,000 − 5,000 included = 45,000 × $0.002$90.00
Total$549.00

If metric points average 0.3 KB instead, metrics are 259.2 GB and the total is $596.76. That compares with $3,056.50 on Datadog's annual prices, or $2,549.50 without the RUM Measure and synthetics lines that Maple does not price the same way. Maple records replays only for the sessions you sample. It has no cheap tier for counting every session the way RUM Measure does: recording all 500,000 sessions would cost $990.00, more than Datadog's $75. Maple includes 10 uptime monitors, which this guide does not compare check for check with the 864,000 API runs, and it has no browser tests. It does not offer database monitoring, security products or Datadog's catalog of integrations, and it keeps logs 30 days.

Datadog is the better fit when you rely on those products, or on its agent's integrations for hosts and managed services. Maple fits teams that already send OpenTelemetry and want a bill that moves with gigabytes rather than hosts and tags.

Maple

Price your own month

The calculator prices Datadog and Maple on the same sliders, and the compare page goes product by product.

Sources and further reading

FAQ

자주 묻는 질문

How much does Datadog cost per host?
Infrastructure Pro is $15 per host per month billed annually or $18 on-demand. Enterprise is $23 or $27. APM adds $31 per APM host billed annually or $36 on-demand, so a host with both costs $46 a month on an annual plan. Logs, custom metrics above the included 100 per host, RUM and synthetics are billed on top (prices verified 2026-11-21).
Is Datadog free?
Datadog has a free tier for up to five hosts with one day of metric retention. It does not include APM, log management or the other paid products, which are billed per host or per unit from the first one.
How does Datadog bill custom metrics?
A custom metric is a metric name plus one unique combination of tag values, including the host tag. Datadog counts the distinct combinations each hour and averages them over the month. Infrastructure Pro includes 100 per host and Enterprise 200, pooled across the account. The list price above that is $5 per 100 custom metrics per month.
What is the difference between ingested and indexed logs in Datadog?
Ingest is $0.10 per GB for every log Datadog receives. Indexing is a second charge per million log events you keep searchable, priced by retention: $1.06 for 3 days, $1.27 for 7, $1.70 for 15 and $2.50 for 30 days on annual billing. Logs you exclude from indexes still pay the ingest charge.
What is the difference between annual and on-demand Datadog pricing?
Annual prices need a commitment. On-demand pays for usage with no commitment, at a higher rate: Infrastructure Pro is $18 instead of $15 per host, APM $36 instead of $31, and 15-day log indexing $2.55 instead of $1.70 per million events. For logs, Datadog says volume above a monthly commitment pays 50% over the annual price.

오늘 첫 트레이스를 확인하세요.

SDK를 추가하고 OTLP 익스포터를 Maple로 향하게 하면 트레이스가 도착합니다.

maple.dev: OpenTelemetry 위의 옵저버빌리티