10 Best APM Tools of 2026

1. 1

   Datadog APM

   Distributed tracing and service performance monitoring sold per APM host, inside Datadog's wider monitoring platform.

   8.8overall

   Distributed tracing and service performance monitoring sold per APM host, inside Datadog's wider monitoring platform.

   Best forTeams already on Datadog infrastructure monitoring

   Adoption data

   4,860Datadog counted about 4,860 customers paying $100,000 or more in annual recurring revenue on September 30, 2026, against about 4,060 a year before.

   Datadog, Datadog Announces Third Quarter 2026 Financial Results (Form 8-K, Exhibit 99.1) (2026): “As of September 30, 2026, we had about 4,860 customers with ARR of $100,000 or more, an increase of 20% from about 4,060 as of September 30, 2025.”

   Datadog APM is the most complete tracing product on this list for teams that want traces, logs and host metrics side by side. The pricing page sets out three tiers: APM from $31 per host per month billed annually, APM Pro from $35 with Data Streams Monitoring, and APM Enterprise from $40 with Continuous Profiler, which records wall, lock, CPU and I/O time at the code level. APM can be bought without Infrastructure Monitoring at $36 per host. Every APM host also gets Universal Service Monitoring, which maps services and dependencies without code changes. Datadog's third quarter 2026 filing reports about 4,860 customers paying $100,000 or more a year, the largest filed figure here. The trade-off is cost: per-host fees apply to every host that sends traces, and span indexing is billed on top. We compared documented features and published prices and did not run our own benchmarks.

   Strengths

   • +Three published APM tiers ($31, $35, $40 per host) make budgets easy to model
   • +Traces link to metrics, logs, processes and network data in one product
   • +Largest filed count of $100,000+ customers among the vendors here

   Limitations

   • −Only hosts sending traces count, but every such host is billed even if traffic is light
   • −Indexed spans and ingest beyond allowances are billed separately
   • −Standalone APM costs $5 more per host than APM bought with infrastructure

   SummaryDatadog APM is the best APM tool for most engineering teams in 2026 because it publishes per-host APM prices from $31 and ties traces to logs, metrics and profiles. Budget for span indexing and every traced host before you commit.

2. 2

   Dynatrace

   A full-stack observability platform whose Full-Stack Monitoring tier bundles APM with host and Kubernetes monitoring.

   8.7overall

   A full-stack observability platform whose Full-Stack Monitoring tier bundles APM with host and Kubernetes monitoring.

   Best forLarge enterprises that want automated root cause analysis

   Adoption data

   $2,207 millionDynatrace's total annual recurring revenue reached $2,207 million in the quarter ended September 30, 2026, a 16% rise on a year earlier.

   Dynatrace, Dynatrace Reports Second Quarter Fiscal Year 2027 Financial Results (Form 8-K, Exhibit 99.1) (2026): “Total ARR of $2,207 million, an increase of 16% on an as reported and constant currency basis”

   Dynatrace sells APM as part of Full-Stack Monitoring, priced at $58 a month per 8 GiB host, which also covers host, process and network analysis. The pricing page adds Kubernetes metrics on Full-Stack hosts at no extra charge, traces at $0.20 per GiB ingested with 10 days of retention, and Code Monitoring for live production debugging at $3.60 per container a month. Its main draw for APM buyers is automated root cause analysis, which links a slow request to the service, process or host causing it. Dynatrace reported $2,207 million of annual recurring revenue for the quarter ended September 30, 2026, up 16 percent. The price unit is the catch: hosts are billed per 8 GiB of memory, so estates with large hosts should model the bill carefully. We compared documented features and published prices rather than testing the product ourselves.

   Strengths

   • +APM and host monitoring come in one per-host price
   • +Kubernetes monitoring is included on Full-Stack hosts at no extra charge
   • +Second largest filed revenue base in this list, at $2,207 million of ARR

   Limitations

   • −Pricing scales with host memory, so large hosts cost more than one unit
   • −Highest entry price for APM among the per-host vendors here
   • −Trace retention beyond 10 days is billed per GiB-day

   SummaryDynatrace is the best APM tool for large enterprises because Full-Stack Monitoring at $58 per 8 GiB host bundles traces, hosts and Kubernetes with automated root cause analysis. Smaller teams will find cheaper entry points elsewhere on this list.

3. 3

   Grafana Cloud

   A hosted Grafana stack whose Application Observability product builds on OpenTelemetry and Prometheus data.

   8.6overall

   A hosted Grafana stack whose Application Observability product builds on OpenTelemetry and Prometheus data.

   Best forTeams standardizing on OpenTelemetry on a budget

   Adoption data

   10,000+Grafana Labs states that more than 10,000 customers, from startups to enterprises, use its products.

   Grafana Labs, About Grafana Labs (2026): “We're trusted by over 10,000+ customers, from startups to enterprises”

   Grafana Cloud's Application Observability product gives teams APM views on top of OpenTelemetry and Prometheus data, so instrumentation is not tied to one vendor's agent. The pricing page lists a permanent free tier with 2,232 host hours and 37,944 container hours a month, roughly three hosts running full time, plus 14 days of retention for metrics, logs, traces and profiles. On Pro, Application Observability costs $0.01 per host hour, about $7.20 per host per month, on top of a $19 monthly platform fee, with automatic volume discounts. Grafana Labs says more than 10,000 customers use its products, a company figure rather than a filed one. The cost model has more moving parts than a single per-host price, because trace and log volume are billed separately. We compared documented features and published prices and did not run our own load tests.

   Strengths

   • +Lowest per-host APM rate in this list at about $7.20 a month
   • +Permanent free tier covers about three hosts running all month
   • +Open standards make it easier to switch backends later

   Limitations

   • −Traces, logs and metrics are billed as separate usage on top of host hours
   • −Free tier support is community only
   • −Customer count comes from Grafana Labs' own site, not an audited filing

   SummaryGrafana Cloud is the best value APM option because Application Observability costs about $7.20 per host on Pro and the free tier covers roughly three hosts. Estimate trace volume as well, since telemetry is billed on top.

4. 4

   Sentry

   An error monitoring and tracing tool built for application developers, priced by event volume rather than hosts.

   8.6overall

   An error monitoring and tracing tool built for application developers, priced by event volume rather than hosts.

   Best forDevelopers who want errors and slow traces in one tool

   Adoption data

   130k+Sentry says more than 130,000 customers use its software monitoring tools.

   Sentry, Sentry customer stories (2026): “Learn why more than 130k customers build better software, faster, with Sentry.”

   Sentry comes at APM from the developer side: it started as error monitoring and now traces requests so a slow endpoint and the exception behind it appear together. The pricing page lists a free Developer plan limited to one user that includes error monitoring and tracing, and a Team plan from $26 a month billed annually with unlimited users, 50k errors and 5M spans. Logs and application metrics include 5 GB, then cost $0.50 per GB. Sentry says it supports more than 100 coding languages and that more than 130,000 customers use it, a company figure. Because pricing follows event volume rather than hosts, it is often cheaper than per-host APM for small teams. It does not replace infrastructure monitoring, so teams that need host and container metrics usually pair it with another tool. We compared documented features and published prices only.

   Strengths

   • +Free plan includes error monitoring and tracing for one user
   • +Team plan has unlimited users at $26 a month
   • +Pricing by events avoids per-host fees for large fleets with little traffic

   Limitations

   • −Free plan is limited to one user
   • −Infrastructure monitoring is not its focus, so host metrics need another tool
   • −Extra logs and metrics beyond 5 GB cost $0.50 per GB

   SummarySentry is the best APM tool for small product teams because the $26 Team plan covers unlimited users, 5M spans and 50k errors. Pair it with an infrastructure tool if you also need host metrics.

5. 5

   Elastic Observability

   APM, logs and metrics on the Elasticsearch engine, sold as a serverless project priced per GB.

   8.4overall

   APM, logs and metrics on the Elasticsearch engine, sold as a serverless project priced per GB.

   Best forTeams that already search logs in Elasticsearch

   Adoption data

   $478 millionElastic posted $478 million of total revenue for the quarter ended July 31, 2026, up 15%, a figure that also includes its search and security products.

   Elastic, Elastic Reports First Quarter Fiscal 2027 Financial Results (Form 8-K, Exhibit 99.1) (2026): “Total revenue was $478 million, an increase of 15% year-over-year, as reported and on a constant currency basis”

   Elastic Observability runs APM on the same Elasticsearch engine many teams already use for logs, so traces and logs live in the same data store. Its serverless pricing page bills by data, not hosts: ingest for logs, traces, synthetics and RUM is as low as $0.07 per GB and retention as low as $0.017 per GB per month, with 50 GB of egress free. That model suits estates with many lightly loaded services, where per-host APM would be costly. Elastic reported $478 million of revenue for the quarter ended July 31, 2026, up 15 percent, though that total includes search and security. The downside is forecasting: trace volume grows with traffic and sampling choices, so bills move with usage. We compared documented features and published prices rather than running our own tests.

   Strengths

   • +No per-host fee, so many small services cost little to trace
   • +Traces sit in the same Elasticsearch store as logs
   • +Public company with filed revenue figures

   Limitations

   • −Bills depend on data volume, which is harder to predict than host counts
   • −Filed revenue covers search and security, not only observability
   • −Usage-based billing has no flat plan price to budget against

   SummaryElastic Observability is the best APM tool for teams already on Elasticsearch because traces land next to logs at ingest prices as low as $0.07 per GB. Set sampling rules early to keep usage costs predictable.

6. 6

   New Relic

   A usage-priced observability platform with APM included, billed on data ingest and full platform users.

   8.3overall

   A usage-priced observability platform with APM included, billed on data ingest and full platform users.

   Best forSmall teams that want full APM free up to 100 GB a month

   New Relic prices APM as part of one platform: you pay for data ingested and for full platform users, not for hosts. The pricing page lists a perpetual free tier with 100 GB of ingest a month, one free full platform user, unlimited basic users and access to more than 50 capabilities, then $0.40 per GB beyond the free allowance. On Standard, the first full platform user costs $10 and each additional one $99, up to five; Pro full users cost $349 each on annual terms. For a small team, that makes full APM with traces and errors cheaper than most per-host tools. The seat model becomes the main cost as teams grow. New Relic is privately held and publishes no current customer count we could cite, so its adoption score is conservative. We compared documented features and published prices and did not run load tests.

   Strengths

   • +APM is included in the free tier rather than sold per host
   • +100 GB a month free is enough for many small production apps
   • +Basic users are free, so viewers do not add cost

   Limitations

   • −Full platform users cost $99 each after the first on Standard
   • −Standard caps full platform users at five
   • −New Relic publishes no customer count we could cite

   SummaryNew Relic is the best free APM option because its free tier includes 100 GB of ingest a month and one full platform user. Larger teams should price the $99 and $349 full-user seats carefully.

7. 7

   Splunk AppDynamics

   Cisco's APM product for hybrid and on-premises applications, linking performance to business transactions.

   7.9overall

   Cisco's APM product for hybrid and on-premises applications, linking performance to business transactions.

   Best forEnterprises monitoring on-premises and three-tier applications

   Adoption data

   $1,095 millionCisco reported $1,095 million of Observability revenue for fiscal 2026, up 4%, a total that includes AppDynamics alongside Splunk Observability and network assurance.

   Cisco Systems, Cisco Systems Form 10-K for the fiscal year ended July 25, 2026 (2026): “Collaboration 4,300 4,154 4,113 146 4 % Observability 1,095 1,055 837 40 4 %”

   Splunk AppDynamics is the APM product Cisco positions for hybrid and on-premises applications, especially traditional three-tier business systems. Its pricing, published on Splunk's observability pricing page, starts with an Infrastructure Edition at $6 per vCPU a month, then Premium Edition from $33 per host and Enterprise Edition from $50 per host, all billed annually. The vendor's pitch is linking application performance to business outcomes, for example which transactions slow down when a database tier degrades. Cisco's fiscal 2026 filing reports $1,095 million of Observability revenue, up 4 percent, but the figure combines AppDynamics with Splunk Observability and network assurance, so its standalone scale is not disclosed. We compared documented features and published prices only.

   Strengths

   • +Covers on-premises and hybrid applications that cloud-first tools handle less well
   • +Published edition prices for an enterprise APM product
   • +Free trial available

   Limitations

   • −Annual billing only at list prices
   • −No standalone adoption figure; Cisco reports it inside a wider category
   • −Less suited to small cloud-native teams than event-priced tools

   SummarySplunk AppDynamics is the best APM tool for on-premises enterprise applications because Premium Edition starts at $33 per host and targets three-tier business systems. Cloud-native teams should look at Splunk Observability Cloud or Datadog instead.

8. 8

   Splunk Observability Cloud

   Cisco's observability suite, with APM and profiling sold in host-based bundles.

   7.9overall

   Cisco's observability suite, with APM and profiling sold in host-based bundles.

   Best forOrganizations already using Splunk for logs or security

   Adoption data

   $1,095 millionCisco booked $1,095 million of Observability revenue in fiscal 2026, up 4%, a category that covers Splunk Observability, AppDynamics and network assurance.

   Cisco Systems, Cisco Systems Form 10-K for the fiscal year ended July 25, 2026 (2026): “Collaboration 4,300 4,154 4,113 146 4 % Observability 1,095 1,055 837 40 4 %”

   Splunk Observability Cloud bundles APM rather than selling it alone. The pricing page lists Infrastructure from $15 per host per month, App & Infra from $60, which adds APM with Always On Profiling and synthetic API monitoring, and End-to-End from $75, which adds real user monitoring and synthetic browser checks. All three are billed annually, and database monitoring is an add-on through sales. Every bundle includes Log Observer Connect, which makes it a natural fit for organizations that already run Splunk. Cisco reported $1,095 million of Observability revenue for fiscal 2026, up 4 percent, but that category also includes AppDynamics and network assurance. At $60 per host for APM, it costs about twice Datadog's entry APM rate. We compared documented features and published prices rather than testing the product ourselves.

   Strengths

   • +Profiling is included with APM rather than sold separately
   • +Fits teams that already keep logs in Splunk
   • +Backed by Cisco's filed Observability revenue

   Limitations

   • −APM requires at least the $60 per host bundle
   • −Prices are annual and sales led beyond the list rates
   • −Cisco's figure mixes several products, so product-level adoption is unclear

   SummarySplunk Observability Cloud is the best APM tool for Splunk customers because the $60 App & Infra bundle includes APM with profiling and every bundle includes Log Observer Connect. Teams without Splunk can find lower per-host rates.

9. 9

   Honeycomb

   An event-based observability tool built for querying high-cardinality traces with unlimited custom fields.

   7.7overall

   An event-based observability tool built for querying high-cardinality traces with unlimited custom fields.

   Best forTeams debugging high-cardinality distributed systems

   Honeycomb treats each traced request as a wide event and lets engineers group and filter by any field, such as user ID, region or feature flag, which suits debugging problems that only affect some requests. The pricing page lists a free plan with up to 20M events and 100M time series data points a month, and Pro from $150 a month for 50M events, scaling up to 750M, with 100 triggers, two service level objectives and single sign-on. Its telemetry pipeline for OpenTelemetry processing starts at $0.10 per GB. Pricing is by event volume with unlimited custom fields, so adding detail to spans does not add cost. Honeycomb publishes no customer count we could cite, which keeps its adoption score low. It is narrower than full-stack suites on infrastructure monitoring. We compared documented features and published prices and did not run benchmarks.

   Strengths

   • +Pricing by events, not hosts or users
   • +Free plan is large enough for real production tracing
   • +Built for slicing traces by any field, such as customer or build ID

   Limitations

   • −No citable customer count, so adoption is scored conservatively
   • −Pro includes only 2 SLOs
   • −Less infrastructure monitoring than full-stack suites

   SummaryHoneycomb is the best APM tool for debugging high-cardinality systems because every event carries unlimited custom fields and Pro starts at $150 for 50M events. Teams that need host monitoring will want a second tool.

10. 10

    IBM Instana

    IBM's automated observability platform, priced per managed virtual server with unlimited users.

    7.5overall

    IBM's automated observability platform, priced per managed virtual server with unlimited users.

    Best forIBM customers and teams that need a self-hosted option

    IBM Instana is the only tool on this list with a published self-hosted price, which matters for teams that cannot send telemetry to a public SaaS. IBM's pricing page lists three plans: pay-per-use from $0.04 per managed virtual server hour with no time commitment, SaaS from $23.10 per MVS a month with unlimited users and no limits on hosts, and self-hosted from $385.20 per MVS a year with feature parity to SaaS. Each MVS can be bundled as Essentials for infrastructure or Standard for full-stack observability, which includes APM. IBM says Instana supports more than 300 technologies. The MVS unit is not the same as a host or GB, so compare quotes with real estate sizes. IBM does not report Instana adoption separately, so we score it conservatively. We compared documented features and published prices only.

    Strengths

    • +Self-hosted edition suits regulated and sovereign environments
    • +Unlimited users and no host caps on SaaS
    • +Pay-per-use plan lets small teams start without a contract

    Limitations

    • −MVS pricing units differ from other vendors, so comparisons take work
    • −IBM publishes no Instana-specific customer figure
    • −Fewer free options than event-priced tools

    SummaryIBM Instana is the best APM tool for regulated teams that must self-host, because a self-hosted plan with SaaS feature parity starts at $385.20 per MVS a year. Others can try it on pay-per-use from $0.04 per MVS hour.

Questions or corrections: editorial@statlyst.com · Disclosure