Datadog pricing (2026): what a small team actually pays

Datadog is free for up to five Basic hosts with 1-day metric retention. Paid infrastructure monitoring starts at $15 per host a month billed annually ($18 on demand), APM adds $31 per host, and logs, custom metrics, containers and tests each have their own meter. Our worked bills for three to eight hosts come to about $160 to $670 a month.

Every number below is from Datadog's own pricing pages and billing docs, checked on November 3, 2026, and cited inline.¹,² Prices are US dollars list price, before any negotiated discount. Fixter is our product; it appears once, near the end, where it is a real alternative.

The host plans at a glance

Datadog does not sell one plan. It sells a catalogue of products, each with its own unit, and infrastructure monitoring per host is the base most accounts start from. You pick one infrastructure level for the account; if it is Pro or Enterprise, you can put a subset of hosts on Basic.²

List prices on November 3, 2026. Datadog describes the free tier as up to five Basic hosts with 1-day metric retention, so we read its limits as those of Basic. DevSecOps Pro ($22) and DevSecOps Enterprise ($34) bundle cloud security with the host plans. Volume discounts start at 500 hosts.

Two things matter for a small team. First, the host price is the floor, not the bill: APM, logs and most other products are separate lines on top. Second, the monitoring products are priced per host or per unit of data, not per user, so team size barely moves the bill. A few products are priced per seat or per committer, such as On-Call at $20 a seat and Incident Management at $30, but none of them is required to monitor an app.¹

Infrastructure Basic, the newest tier, launched on October 28, 2026. Datadog describes it as baseline visibility for noncritical workloads: CPU, memory, disk, network, uptime and system load, out-of-the-box dashboards, and monitors that can notify the tools you already use. The pricing list shows it as contact sales on every billing option.¹,⁴

The meters a small team actually hits

Datadog's pricing list has more than a hundred line items.¹ A 1 to 10 developer SaaS with a backend, a database and real users touches about ten of them. These are those ten, at all three billing options.

From Datadog's pricing list on November 3, 2026. Starred items on that list (containers, custom metrics, APM ingestion and indexed spans) note that existing customers should check their own billing agreement for included allotments.

What each one means in practice:

• APM is a second host fee. Every host sending traces is an APM host, including an OpenTelemetry Collector. With the infrastructure pricing model, every APM host also needs Pro or Enterprise infrastructure, so a traced host costs $46 a month billed annually, not $15. Each APM host includes 150 GB of ingested spans and 1 million indexed spans a month, pooled across hosts.²
• Logs are billed twice. Ingesting a log costs $0.10 per GB; making it searchable and usable in monitors costs $1.70 per million events for 15 days, billed annually. Logs you only archive cost the ingest. Flex Logs storage, for logs you search rarely, starts at $0.05 per million events stored but does not support monitors.²
• Custom metrics are where bills surprise people. Pro includes 100 per host, pooled and averaged across the account. Log-based metrics count against the same allotment. A custom metric is one metric name with one combination of tag values, host included, so a single metric tagged by customer or endpoint can be hundreds.²,⁶
• Containers. Pro includes 5 per host and Enterprise 10. Extra containers are $0.002 per container hour on demand, about $1.46 a month for one that runs all month, or $1 a month prepaid.²
• Bits AI investigations. Datadog's AI features draw from AI credits. Its estimate for one autonomous investigation is about 4 credits, with the caveat that actual use varies with the task. At $1.30 a credit on demand that is about $5.20 an investigation; the annual bundle is $500 for 500 credits a month, and unused credits expire at the end of the month.²

How Datadog counts what you pay for

Three billing rules decide whether a Datadog bill matches the estimate you made.

1. Hosts are counted hourly and billed on the high-water mark. Datadog meters hosts every hour, drops the top 1 percent of hours, and bills the highest remaining count. In a 720 hour month that ignores about seven hours of a spike. An autoscaling group that stays doubled for a day is billed doubled for the month.³
2. Custom metrics are averaged. They are counted hourly and averaged across the month and across the whole account, not per host. One deploy that adds a high cardinality tag for a week raises the average for that month.²,³
3. Committed logs have a 50 percent overage. If you commit to a monthly log volume and go over, Datadog keeps the data and charges the excess at the annual price plus 50 percent. On demand, the default index retention is 15 days.²

The billing cycle starts on the first of the month in UTC, and a first month is prorated.³ Annual billing is the cheapest column but is a commitment; month to month and on demand cost about 20 percent more on the host lines and up to 50 percent more on indexed logs.¹

Three worked bills for a 1 to 10 developer team

These are list prices multiplied out from the tables above. The usage numbers are our assumptions for a typical small SaaS; your own will differ, but the shape will not.

One founder, two servers, host health only

Free tier, $0. You get core system metrics for up to five Basic hosts and alerts on them. The limits that bite: metrics are kept for one day, the free tier is infrastructure only (APM and log indexing are paid lines), and container monitoring is not available on Basic hosts.² It answers "is the box up and is the disk full", not "why did checkout fail".

Two developers, three hosts, traces and logs

Three Pro hosts ($45) with APM on all three ($93). 20 GB of logs a month ingested ($2) and 10 million of those events indexed for 15 days ($17). About $157 a month billed annually, or about $190 on demand ($54, $108, $2 and $25.50). Everything else stays inside the allotments: 300 custom metrics and 450 GB of span ingestion for the account.

Six developers, an eight node Kubernetes cluster

Eight Pro hosts ($120), 60 containers of which 40 are included, so 20 prepaid ($20). APM on all eight nodes ($248). 300 GB of logs ingested ($30), 50 million events indexed for 15 days ($85). 1,200 custom metrics against 800 included, so 400 extra ($20). Five API checks every five minutes, 43,200 runs ($21.60). About $545 a month billed annually, about $670 on demand. Add 20 Bits AI investigations on demand, about 80 credits, and it is roughly $650 annually.

The pattern: for a small team the host price is the visible part, and APM roughly triples it. After that, indexed logs and custom metrics are the lines that move month to month, because they grow with traffic and with how your code logs and tags, not with anything you bought on purpose.

When a cheaper setup is enough

First, when Datadog is the right buy. If you have a platform team, many hosts already instrumented, a need for its 1,000+ integrations, or security and compliance products in the same place, Datadog is hard to beat and the bill is the price of that breadth. The same holds if your team already knows it well. The advice below is for the other case: a 1 to 10 developer team that needs to know when production breaks and why.

Our judgement, not Datadog's. Prices and limits as in the tables above.

• Stay on the free tier while it answers your question. Five hosts of core metrics is a reasonable start for a single founder. Know that it will not tell you why a request failed.
• Index on purpose. Exclude debug and health check logs from the index, keep 7-day retention where 15 is not needed, and use log-based metrics sparingly, since they eat the custom metric allotment.²
• Watch tag cardinality. Never tag a custom metric with a user id, request id or anything unbounded.
• Use the startup program if you qualify. Our Datadog for startups guide covers how the credits work and what happens when they run out.
• Compare the whole setup, not the host price. Our best monitoring tools for a small SaaS comparison lists the main options with setup, free tier and price at small scale, and Datadog alternatives goes deeper on each.

Where Fixter is the cheaper setup

Fixter (fixter.dev) is our product, so weigh this with that in mind. Fixter finds and fixes the issues in your app in production, from your coding agent: errors, failing infrastructure, slow requests and downtime. It finds the root cause and fixes it: in your agent over MCP, or as a pull request ready for your review. Nothing to learn, built for small teams.

We do not claim to do monitoring better than Datadog. The difference is the shape. There is no per-host or per-product meter: you pay for the data you send over standard OpenTelemetry, with unlimited users on every plan. There is a Free plan, and paid plans start at €20 a month.⁵ And there is no console to learn first. Setup is one prompt in your coding agent, "Read fixter.dev/llms.txt and set up Fixter.", and you ask about production from the same session, in plain language. What we claim is the time and knowledge it takes to go from nothing to a useful answer about your app.

Where Datadog is still the better buy: you need its integrations catalogue, its security products, or deep infrastructure and Kubernetes views, or you already run it across many hosts. Fixter is a younger product with a narrower scope. If you want your coding agent to read the Datadog data you already have, our Datadog MCP vs Sentry MCP post covers what Datadog's MCP server can do.

Frequently asked questions

Is Datadog free?

Partly. Datadog's free tier covers infrastructure monitoring for up to five Basic hosts with 1-day metric retention, at no cost. APM, log indexing, custom metrics beyond the host allotment, synthetic tests and the other products on the pricing list are billed per unit. To monitor more hosts or keep data longer you move to a paid plan.

How much does Datadog cost per host?

Infrastructure Pro is $15 per host a month billed annually and $18 month to month or on demand. Infrastructure Enterprise is $23 annually and $27 otherwise. APM is a second per-host charge: $31 a month billed annually with Infrastructure Monitoring attached, $36 on demand, or $36 standalone. Infrastructure Basic is priced through sales.

What is Datadog Infrastructure Basic?

A host tier Datadog announced on October 28, 2026 for noncritical workloads. It collects core system metrics (CPU, memory, disk, network, uptime, load), ships out-of-the-box dashboards and can be mixed with Pro or Enterprise hosts in the same account. Container monitoring is not available on Basic hosts and it has no custom metric allotment. The price is contact sales.

How does Datadog count hosts for billing?

Datadog meters hosts every hour. At the end of the month it bills the highest hourly count after dropping the top 1 percent of hours, which Datadog calls the high-water mark of the lower 99 percent. A short autoscaling spike is ignored; a fleet that stays bigger for more than about seven hours a month is billed at the bigger size.

How much do Datadog logs cost?

Logs are billed twice. Ingesting costs $0.10 per GB. Indexing, which makes logs searchable and usable in monitors, is priced per million log events by retention: $1.70 per million for 15 days billed annually and $2.55 on demand. Flex Logs storage for logs you rarely search starts at $0.05 per million events stored.

How much does a Datadog Bits AI investigation cost?

Bits AI runs on AI credits. Datadog estimates about 4 credits per autonomous investigation, and says actual use varies. Credits cost $500 for a 500 credit bundle a month on an annual or monthly plan, or $1.30 per credit on demand, so one investigation is roughly $4 to $5 at list price. Unused credits do not roll over.