Datadog pricing explained: hosts, logs, traces, and a worked example

Estimating a Datadog bill means accounting for several billing units. Infrastructure and APM have host-based charges; logs have ingestion and indexing charges; metrics, replays, synthetics, and other products add their own usage.

This article works through one example with the assumptions shown, then explains how to make a comparable estimate for Moneat.

Prices below come from Datadog’s published price list and were reviewed on September 13, 2026. They use the billed annually column unless noted. Contracts, included allotments, discounts, commitments, taxes, and product configuration can change the final invoice.

The published billing units

Datadog publishes separate month-to-month and on-demand columns. For example, Infrastructure Pro is listed at $18 per host month-to-month, APM at $36, 30-day indexed logs at $3 per million events, and synthetic API tests at $6 per 10,000 runs. Always match the column to the contract you are modeling.

Correcting the ten-million-log-events example

Ten million indexed log events per day over a 30-day month is 300 million events:

10 million events/day × 30 days = 300 million events

15-day retention: 300 × $1.70 = $510/month
30-day retention: 300 × $2.50 = $750/month

That is the indexing charge at the listed annual-billing rate, not $5,100. Log ingestion is separate and cannot be calculated from the event count alone; you also need the number of ingested gigabytes.

A worked monthly example

Consider a SaaS application with these assumptions:

• 20 hosts on Infrastructure Pro;
• APM enabled on 15 of those hosts;
• 150 GB of logs ingested during the month;
• 150 million log events indexed with 30-day retention; and
• 20 API tests running once per minute for a 30-day month.

The published-list calculation is:

This is an illustration, not a quote. It excludes indexed-span and excess APM-ingestion charges, session replay, RUM, profiling, database and network monitoring, security products, support, and negotiated terms. Datadog also includes product-specific allotments that may change which usage is billable for a particular account.

Why a bill can move unexpectedly

The model becomes difficult when the billing dimensions do not move together:

• Auto-scaling changes the number of billable hosts or containers.
• A logging change can increase both ingested bytes and indexed event count.
• A high-cardinality metric creates more custom metric series than its name suggests.
• Trace sampling and retention determine ingested and indexed span usage separately.
• A shorter synthetic interval multiplies test runs across every location.

Datadog currently lists custom metrics at $5 per 100 custom metrics per month. It separately lists ingested custom metrics at $0.10 per 100. When investigating a metric spike, confirm which unit the invoice uses before multiplying a rate.

Compare the same workload in Moneat

Moneat’s billing model is different, but it is not accurate to collapse every signal into one unlimited bucket. The live model includes a plan price and an ingestion allowance, plus distinct allowances or overages for areas such as APM spans, custom metrics, infrastructure metric series-hours, and analytics pageviews.

Use the pricing calculator with the same inputs you used for Datadog. Include all telemetry, not just logs, and check the current pricing page for the plan and rates the billing service is actually returning.

A fair comparison states:

1. the number and type of hosts;
2. ingested bytes and retained or indexed events;
3. spans, custom metrics, and infrastructure series-hours;
4. RUM, replay, analytics, and synthetic usage;
5. retention and sampling choices; and
6. annual, monthly, on-demand, or negotiated terms.

If one of those inputs is missing, the result is a partial estimate—not evidence that one platform is a fixed multiple cheaper than another.

Model your workload | Review Moneat pricing