New Relic pricing, explained

01 · Pricing model

The New Relic pricing model

People are billed by user type, and only two of the three types cost money. The plan, called an edition (Free, Standard, Pro or Enterprise), sets the price of a full platform user, how many you can add and the support level.

The three user types

Every person in the account has one user type. It decides which screens they can open and what they cost; roles still decide what they may change.

Hosts are not counted at all. A fleet of 10 hosts and a fleet of 1,000 pay the same user fees and differ only in the GB they send. GB ingested is measured after your drop rules run and is rounded up to the next whole GB each month.

Smaller charges sit on top: synthetic checks past your edition's monthly allowance cost $0.005 each, storing data in the EU region adds $0.05 per GB, and extended retention and Advanced Compute are add-ons. Pro and Enterprise accounts can also swap user fees for compute pricing, covered under compute pricing.

02 · Editions

New Relic plans: Free, Standard, Pro and Enterprise

Every edition includes unlimited basic users and the same free 100 GB a month. They differ in the price and number of full platform users, the synthetic checks included, and support: Standard gets ticket support, Enterprise adds FedRAMP and HIPAA eligibility and custom terms. Prices below are what you pay each month.

Standard's five-user limit

Standard allows five full platform users. A sixth person who needs the APM views moves the account to Pro, where every full platform user costs $349 a month (with a 1-year commitment) instead of $99 a month. Five Standard users cost $406 a month; six Pro users cost $2,094 a month with the commitment, or $2,513 a month without.

The limit counts full platform users only. Core and basic users are unlimited on Standard, which is why the cheaper option in the cost example stays on Standard with 12 people.

Pay-as-you-go or a commitment

Standard and Pro can run pay-as-you-go: New Relic bills at the end of each calendar month, with no commitment past it. On Pro, committing for a year lowers the full platform price from $418.80 to $349 a month, 17% less; you still pay monthly. Pro and Enterprise customers can also sign a commitment contract for volume discounts. New Relic runs programs for startups and nonprofits and does not publish their discounts.

03 · Free tier

The New Relic free tier

The free tier is permanent and needs no credit card. Each month it includes:

Past 100 GB

The same 100 GB allowance behaves differently on the free tier and on a paid plan once a month's data runs past it:

One calendar month of ingest. The allowance resets on the first day of each month.

On the free tier, passing 100 GB stops ingest and locks you out of the platform until you upgrade or the month ends. A burst of debug logs in the third week can leave you without data or dashboards for the rest of the month. Alert on your own usage well before New Relic's 85% email; the query is under measure ingest and seats.

Is there a free trial?

New Relic does not advertise a time-limited trial of Standard or Pro. The free tier takes its place: its one full platform user sees the same curated views a paid user does, on up to 100 GB a month, with no end date.

04 · Rates

User types and data rates

How users are counted

Original Data or Data Plus

Each account picks one data option, and both bill from the 101st GB.

Data Plus costs 50% more per GB, so it pays off when you need the retention or the compliance eligibility. Some data never counts toward GB ingested: basic alerting data, aggregated infrastructure data and dimensional metrics, golden metrics, synthetic check results, and New Relic's own usage and audit events.

Compute pricing (CCUs)

Pro and Enterprise can replace user fees with Core Compute. You still pay for data, plus compute capacity units (CCUs): the capacity each successful action you run consumes, such as loading a page or running a query. Every user gets full platform access. Requests that fail on New Relic's side and queries that analyze your own usage are free. New Relic does not publish a per-CCU price, and its billing docs list Core Compute as a public preview arranged through an account executive.

Advanced Compute is a separate add-on, billed in advanced CCUs, for New Relic's AI, automation and security features. It applies on either model, and basic users can run up charges with it too.

05 · Worked example

New Relic cost: a worked monthly example

A team of 12 engineers sends 800 GB of logs, traces and metrics a month to the US region on Original Data. Hosts cost nothing, so the bill comes from the users and the 700 GB past the free 100. Two ways to staff the same team:

Option B stays on Standard because the five-user limit counts only full platform users. It works if the eight core users can do without the curated APM and infrastructure views, which usually means the people who are not on call.

How the total moves

For comparison, the same 800 GB on Maple's Startup plan comes to about $189 a month with no per-user fee: $39 including 100 GB per signal, then $0.30 per GB, assuming the volume splits evenly across logs, traces and metrics. The Maple vs New Relic comparison has a calculator for your own numbers.

Dynatrace vs New Relic pricing

The two vendors bill different things. New Relic counts people and GB and ignores hosts. Dynatrace counts hosts and GiB of logs and traces, charges nothing per user, and sells through an annual minimum commitment.

Before data, five Standard users on New Relic cost $406 a month, while 100 full-stack hosts of 8 GiB on Dynatrace cost $5,800. Swap the sizes and the order flips: 40 Pro users cost $13,960, and 20 Dynatrace hosts cost $1,160. The Dynatrace pricing guide covers its rate card in detail.

06 · Cost control

Lowering a New Relic bill

Seats move the bill more than data does, so start with who holds which user type, then cut data you never query.

1. 01

   Measure ingest and seats

   New Relic records its own usage in the NrConsumption and NrMTDConsumption event types, which do not count toward ingest. Run these in the New Relic query builder. Daily GB by source over three months:

   FROM NrConsumption SELECT sum(GigabytesIngested)
   WHERE productLine = 'DataPlatform'
   SINCE 3 months ago FACET usageMetric TIMESERIES 1 day

   Billable users so far this month:

   FROM NrMTDConsumption SELECT latest(FullPlatformUsersBillable),
     latest(CoreUsersBillable) SINCE 3 hours ago

   GB ingested so far this month. Put an alert condition on it at your budget, or at 70 GB on the free tier:

   FROM NrMTDConsumption SELECT latest(GigabytesIngested)
   WHERE productLine = 'DataPlatform'

2. 02

   Move people to cheaper user types

   Downgrade people who only read dashboards to basic, and people who mostly work in logs and errors to core. Because a month bills at the highest type held, make changes on the first of the month, and on a commitment contract avoid the second downgrade-upgrade cycle.

3. 03

   Drop data before it is billed

   Ingest is measured after New Relic's processing, so dropped data is never billed. Since August 31, 2026, Pipeline Control cloud rules have replaced NRQL drop rules for this inside New Relic, and its gateway can filter in your own Kubernetes cluster. Dropped data cannot be recovered.

4. 04

   Commit once usage is stable

   A 1-year Pro commitment saves $69.80 per full platform user a month, and commitment contracts add volume discounts. The commitment is owed whether you use it or not, so sign after a few months of measured usage.

Filter in an OpenTelemetry Collector

If your services already export OpenTelemetry, an OpenTelemetry Collector running in your infrastructure can drop data before it reaches any vendor. This configuration receives OTLP from your services on the Collector's default ports (4317 for gRPC, 4318 for HTTP), drops DEBUG and TRACE logs (logs with no severity set pass through), and exports traces and logs to New Relic's OTLP endpoint with your license key:

receivers:
  otlp:
    protocols:
      grpc:
      http:

processors:
  filter/drop-debug-logs:
    error_mode: ignore
    logs:
      log_record:
        - severity_number != SEVERITY_NUMBER_UNSPECIFIED and severity_number < SEVERITY_NUMBER_INFO
  batch:

exporters:
  otlp_http/newrelic:
    endpoint: https://otlp.nr-data.net      # EU: https://otlp.eu01.nr-data.net
    headers:
      api-key: ${env:NEW_RELIC_LICENSE_KEY}
  otlp_http/maple:
    endpoint: https://ingest.maple.dev      # EU: https://ingest.eu.maple.dev
    headers:
      Authorization: Bearer ${env:MAPLE_INGEST_KEY}

service:
  pipelines:
    traces:
      receivers: [otlp]
      processors: [batch]
      exporters: [otlp_http/newrelic, otlp_http/maple]
    logs:
      receivers: [otlp]
      processors: [filter/drop-debug-logs, batch]
      exporters: [otlp_http/newrelic, otlp_http/maple]

The otlp_http/maple exporter sends a copy of the same data to Maple, so you can compare what each vendor would bill for identical traffic. Delete it, and its name in both exporters lists, if you only want the filter. Sampling traces and dropping attributes nobody queries cut ingest the same way; sampling and throughput covers how Maple keeps request rates accurate under sampling.

After a few days of dual export, compare the GB each signal sent to Maple with the NrConsumption numbers from the first query.