Datadog vs New Relic: Per-Host Pricing Against Per-User and Per-GB Pricing
Prices and facts checked on vendor pages on Nov 8, 2026.
Datadog and New Relic both bundle APM, infrastructure, logs, browser monitoring, and synthetics, and both now sell AI investigation. The difference that drives most buying decisions is the meter. Datadog charges per host and per product usage. New Relic charges per engineer who needs full access and per gigabyte ingested, and leaves hosts unmetered.
The Short Answer
Pick New Relic if you have a small team and a large or fast-changing fleet, and you want to start on a free tier with one pool of data pricing. Pick Datadog if many people need access, you prefer to buy products one by one, or you want finer control over how long logs stay searchable. Run both numbers against your own headcount, host count, and daily data volume before deciding, because the winner flips depending on which of those is large.
Datadog vs New Relic at a Glance
How Datadog Charges
Rates are annual list prices checked Nov 8, 2026. Monthly and on-demand columns cost more, for example $18 for Infrastructure Pro and $36 for APM.
- Infrastructure Pro is $15 per host per month and includes 100 custom metrics per host. APM is $31 per APM host and includes 150 GB of ingested spans and 1 million indexed spans, summed across your APM hosts.
- Logs are $0.10 per GB ingested. Indexed events are $1.06, $1.27, $1.70, or $2.50 per million for 3, 7, 15, or 30 days.
- Flex Logs Storage is $0.05 per million events, with query compute sold as a separate tier.
- Rehydrating archived logs costs $0.10 per compressed GB scanned, plus your indexing rate for the events it brings back.
- AI Credits are $500 per 500 credits a month on an annual bundle, or $1.30 per credit on demand.
How New Relic Charges
New Relic prices two things and then optional add-ons. Checked Nov 8, 2026:
- Data: 100 GB free each month, then $0.40 per GB. Data Plus is $0.60 per GB and adds up to 90 extra days of retention, a higher query limit, and advanced log obfuscation. EU storage adds $0.05 per GB.
- Users: basic users are free. Core users are $49. Full platform users on Standard are $10 for the first and $99 for each additional, up to five. On Pro they are $349 each on an annual commitment or $418.80 on monthly pay-as-you-go.
- Counting: you pay for data after trimming and transformation rules, and synthetic check data, aggregated infrastructure data, and golden metrics do not count as GB ingested. Extra synthetic checks cost $0.005 each.
- Compute option: New Relic also sells a compute-based model that drops user licenses, available by contacting sales.
Two Worked Scenarios
Both use the list prices above. Assume 25 hosts and 500 GB a month of telemetry, of which 300 GB is logs at roughly 1 KB per event (300 million events), kept 15 days. For Datadog we assume the other 200 GB is covered by host prices and APM allotments, with no extra custom metrics.
Scenario A: 5 engineers need full access.
- Datadog: 25 × ($15 + $31) = $1,150 for hosts, plus $30 ingest and 300 × $1.70 = $510 to index every log event, for $1,690 a month. Indexing only 60 million events brings the log part to $132 and the total to $1,282.
- New Relic Standard: $10 + 4 × $99 = $406 for users, plus 400 GB beyond the free 100 at $0.40 = $160, for $566. On Pro with 5 users, 5 × $349 = $1,745 plus $160 is $1,905.
Scenario B: 40 engineers need full access, same fleet and data.
- Datadog stays at $1,690, because the list has no user fee on these products.
- New Relic Pro: 40 × $349 = $13,960 plus $160 is $14,120. Moving most of those people to Core ($49) or basic users cuts that sharply but removes access to the curated APM and infrastructure screens.
The first scenario favors New Relic, and the second favors Datadog. Neither includes discounts, commitments, or Datadog custom metrics beyond the included 100 per host.
Logs, Traces, and Retention
New Relic keeps one data model. APM agents turn on logs in context and log forwarding by default, so logs appear in APM, infrastructure, tracing, and the errors inbox. You pay the same per-GB rate whether the byte is a log line or a metric, and data that you trim before it is stored is not billed.
Datadog gives you more dials and asks you to set them. Correlating logs and traces depends on trace and span IDs injected into logs that are JSON or parsed into attributes. Datadog also notes that logs and traces are sampled independently, so a log can point to a trace that was never retained. In return you can choose per index how long logs stay searchable, forward every ingested log to your own storage, or use Flex Logs for long retention.
For long-term storage the two differ. New Relic extends logs to seven years through Live Archives, charged separately for storage and queries, and says it does not charge for data egress or hydration. Datadog charges $0.10 per compressed GB scanned when you rehydrate from your own archive.
OpenTelemetry, Query Languages, and AI
New Relic accepts OTLP natively and recommends OTLP over HTTP. Datadog takes OTLP through the DDOT Collector in its Agent, a standard Collector, or direct ingestion, and a Collector with two exporters can feed both. New Relic counts the data that ends up stored after your trimming rules, while Datadog meters ingestion and indexing as separate steps.
NRQL is a SQL-like language that powers New Relic's charts, dashboards, and alerts, so a query written once can be reused. Datadog's Log Explorer works through facets and search syntax.
On AI, Datadog's Bits Investigation forms hypotheses and gathers telemetry to pinpoint root causes, billed in AI Credits at about 4 credits per autonomous run. New Relic's Autopilot investigates incidents, can be mentioned in a Slack channel to triage alerts, and is not yet available for FedRAMP.
When to Pick Datadog
- Many engineers, support staff, and managers need full access and you do not want to pay per seat.
- You want to tune which logs are indexed, which are archived to your own storage, and how long each stays searchable.
- You are fine buying products one at a time and want a free tier based on hosts.
- You need one of its Japan, Australia, or UK sites.
When to Pick New Relic
- You have a small team and many hosts, containers, or short-lived workloads you do not want to count.
- You want a free tier with 100 GB of ingest and one full platform user, and can accept a hard stop if you pass it.
- You want logs to appear inside APM and infrastructure screens without extra correlation work.
- You expect to keep logs for years and prefer a separate archive tier.
Where Sazabi Fits
Sazabi is AI-native observability and security that can run beside either one. Add an otlphttp exporter to your OpenTelemetry Collector to send Sazabi a copy of your logs, or change the New Relic infrastructure agent's logging endpoint, which sends those logs to Sazabi instead of New Relic. Its agent investigates in Slack and opens issues. Sazabi does not offer synthetic checks, so keep your synthetic monitoring tool for those. For one-on-one comparisons, see Sazabi vs Datadog and Sazabi vs New Relic, the Datadog alternatives and New Relic alternatives, and Sazabi pricing.
Related head-to-heads: Datadog vs Dynatrace, Datadog vs Splunk, and Splunk vs New Relic.
Frequently Asked Questions
Is New Relic cheaper than Datadog?
It depends on how many people need full access versus how many hosts you run. New Relic bills users and gigabytes ingested, with 100 GB free and $0.40 per GB after that, and does not charge per host. Datadog bills per host and per usage meter, with no user fee on infrastructure or APM. A small team with many hosts tends to look better on New Relic list prices, while a large team on a modest fleet tends to look better on Datadog. List prices checked Nov 8, 2026.
What happens when I pass the New Relic free tier?
New Relic says data ingest ceases and platform access is lost when you exceed the free 100 GB in a month, until you upgrade or the month rolls over. You get an email and an on-screen alert at 85 percent. The Datadog free tier works differently: it covers up to five Basic hosts with 1-day metric retention rather than a data volume.
How long do New Relic and Datadog keep logs?
New Relic keeps logs for 30 days by default and can extend to seven years with Live Archives, charged separately for storage and queries. Datadog indexes logs for 3, 7, 15, or 30 days, offers Flex Logs for up to 15 months, and can archive all ingested logs to your own cloud storage for rehydration at $0.10 per compressed GB scanned.
Can I send OpenTelemetry data to New Relic and Datadog at the same time?
Yes, through an OpenTelemetry Collector with two exporters. New Relic accepts OTLP natively and recommends OTLP over HTTP, and Datadog accepts OTLP through the DDOT Collector, a standard Collector, or direct OTLP ingestion.
Which one needs less setup to link logs to traces?
New Relic APM agents enable logs in context with log forwarding by default. Datadog correlates logs and traces by injecting trace and span IDs into logs, which must be JSON or parsed into attributes, and it notes that logs and traces are sampled independently, so a log can reference a trace that was not retained.
Sources
- Infrastructure Pro, APM, log ingestion, indexing, and rehydration, Flex Logs Storage, and AI Credit list prices (accessed Nov 8, 2026)
- Free tier of five Basic hosts with 1-day metric retention, free trial, APM allotments per host, custom metric allotments, and billing plan options (accessed Nov 8, 2026)
- Trace and span ID injection for log correlation, JSON or parsed logs, and independent sampling of logs and traces (accessed Nov 8, 2026)
- Flex Logs for long retention with separate query compute (accessed Nov 8, 2026)
- Forwarding all ingested logs to your own cloud storage, with rehydration and archive search (accessed Nov 8, 2026)
- DDOT Collector, OpenTelemetry Collector export, and direct OTLP ingestion into Datadog (accessed Nov 8, 2026)
- Datadog sites are independent and located in the US, EU, Japan, Australia, and the UK (accessed Nov 8, 2026)
- Bits Investigation forms hypotheses, gathers telemetry, and pinpoints root causes (accessed Nov 8, 2026)
- Free tier contents, $0.40 and $0.60 per GB data options, user prices by edition, EU data center surcharge, synthetic check price, and the statement that hosts are not charged (accessed Nov 8, 2026)
- What counts as GB ingested, default 8-day retention, Data Plus 90-day extension, and user type definitions (accessed Nov 8, 2026)
- Retention settings by data type, extension up to seven years, and the Data Plus extension (accessed Nov 8, 2026)
- 30-day default log retention and Live Archives extension up to seven years, with separate storage and query charges (accessed Nov 8, 2026)
- Logs in context across APM, infrastructure, tracing, and errors, enabled by default in APM agents (accessed Nov 8, 2026)
- Native OTLP ingest with US, EU, and JP endpoints (accessed Nov 8, 2026)
- NRQL is a SQL-like query language that also powers charts and dashboards (accessed Nov 8, 2026)
- Autopilot investigates incidents and can be used in Slack, with a HIPAA note and no FedRAMP availability (accessed Nov 8, 2026)
- Sazabi plan prices, included usage, and per-GB overage rates (accessed Nov 8, 2026)
- Pointing the New Relic infrastructure agent logging endpoint at Sazabi, which replaces rather than adds (accessed Nov 8, 2026)
- Adding a Sazabi OTLP exporter beside existing Collector exporters (accessed Nov 8, 2026)