Dynatrace Alternatives
Six Dynatrace alternatives for APM and observability, compared on the unit that makes each bill grow, the commitment each one asks for and what happens when you go over.
Dynatrace sells its platform through an annual spend commitment, and Full-Stack Monitoring runs $58 a month per 8 GiB host, so the bill follows your servers' memory, not just their number. Our picks: Datadog if you want the broadest platform and can watch per-host add-ons, New Relic if your team is small and your data is large, and Grafana Cloud if you want open standards with a free tier. Elastic, IBM Instana and SigNoz cover log-heavy teams, auto-discovery fans and open-source budgets.
The ranking is editorial: a sponsored placement is always flagged on its card, and no affiliate link changes a score.
Why people leave Dynatrace
Dynatrace earns its reputation: one agent, OneAgent, discovers hosts, processes and services, traces requests end to end and feeds an AI engine that points at the root cause of an incident. Few teams leave because the product failed them. They leave because of how it is sold and how the bill moves, and, in smaller teams, because of the time it takes to master. Here are the reasons that come back most often, each with the number behind it.
An annual commitment before anything else
The bill follows memory, not hosts
Many meters on one invoice
Users describe the pricing as hard to predict
A steep learning curve for small teams
Instrumentation tied to OneAgent
Datadog
Datadog is the name nearly every comparison guide puts next to Dynatrace, and a close match in breadth: infrastructure monitoring, APM, logs, real user monitoring, synthetics and security in one interface, with more than 1,000 integrations. Its agent runs on the host or in a container next to your workloads, and it supports OpenTelemetry on the Pro plan. Where Dynatrace leans on automatic discovery, Datadog leans on its integration catalog and its dashboards: you wire more by hand, and you see more of what you wired.
The pricing logic is the opposite of Dynatrace. Dynatrace sells one commitment you spend across capabilities; Datadog sells each product on its own meter. A host is any physical or virtual OS instance you monitor. Infrastructure Pro costs $15 per host a month billed yearly, or $18 on demand. APM adds $31 per host when infrastructure is attached (billed yearly) or $36 on demand. So a host with both runs $46 a month on a yearly plan and $54 on demand: $552 against $648 a year, a $96 gap per host. Multi-year and volume discounts are listed from 500 hosts a month. Logs have their own meters, $0.10 per ingested GB plus $1.70 per million events indexed for 15 days on a yearly plan. Alerts can also trigger workflows outside the tool, as our Datadog and Make integration guide shows, and our guide to leaving Datadog shows what people run into after two years.
- 1,000+ integrations across clouds, databases and queues
- 150 GB of ingested spans per APM host each month
- Free tier for five hosts, useful for a pilot
- ✓Products from infrastructure and APM to RUM, synthetics and security
- ✓Annual, monthly and hourly billing options
- ✓APM can be bought standalone, without infrastructure monitoring
- ✗Each product adds its own per-host or per-GB line
- ✗Log volume above a commitment costs the annual price plus 50%
- ✗Custom metrics beyond the allotment are billed on usage
| Criterion | Datadog | Dynatrace |
|---|---|---|
| Billing unit | Per host, per product, plus GB and events | Memory-GiB-hours, pods, GiB and sessions |
| Commitment | Annual, monthly or hourly | Minimum annual spend |
| APM entry | $31 per host with infrastructure, yearly | $58 per 8 GiB host, Full-Stack |
Pick Datadog if you want a Dynatrace-class platform with a free pilot and you can track each product's meter. Take New Relic instead if most of your cost is data volume and only a handful of engineers need full access.
New Relic
New Relic is the other name nearly every guide ranks, and it answers the Dynatrace pricing complaint with a different meter: you pay for data and for users, not for hosts. Every account gets 100 GB of free data ingest a month, and the free tier adds one free full platform user, for as long as it exists. Watch the cap: past 100 GB in a month, a free account stops ingesting and loses platform access until it upgrades or the next month starts. Beyond that, standard data costs $0.40 per GB, or $0.60 per GB with Data Plus, which adds up to 90 days of extra retention and, on Enterprise, FedRAMP Moderate and HIPAA eligibility. Its APM, infrastructure, logs, browser and synthetics all draw from the same pool, so turning on a core capability does not add a SKU; only Advanced Compute features such as Transaction 360 and New Relic AI bill compute units on top.
The catch sits in the user line. On Standard, the first full user costs $10 and each additional one $99, up to a maximum of five. A team of five pays $406 a month in seats. Past five, you move to Pro, where a full user costs $349 a month on an annual commitment or $418.80 on monthly pay as you go. Ten engineers on Pro cost $3,490 a month yearly or $4,188 monthly: a $698 monthly gap, $8,376 a year, just for avoiding the commitment. Core users, who need less than the full platform, cost $49. Pro and Enterprise can also move to compute-based pricing with no user licenses, priced on request. The data math is friendlier: 500 GB in a month means 400 billable GB, or $160 at the standard rate and $240 with Data Plus. Our New Relic switch guide covers the reverse move.
- 100 GB of data free every month, no time limit
- One price per GB across APM, logs, infrastructure and browser
- Core users at $49 for people who mostly read
- ✓No per-host fees, so adding hosts adds no host line
- ✓A real free tier, not a trial
- ✓Data Plus adds up to 90 days of extra retention for $0.20 more per GB
- ✗Standard stops at five full platform users
- ✗Pro users cost $418.80 a month without an annual commitment
- ✗The free tier cuts ingest and access above 100 GB in a month
| Criterion | New Relic | Dynatrace |
|---|---|---|
| Billing unit | GB ingested plus users | Memory-GiB-hours, pods, GiB and sessions |
| Free tier | 100 GB a month, 1 full user | 15-day trial |
| Commitment | Optional, monthly pay as you go available | Minimum annual spend |
Pick New Relic if fewer than six engineers need full access and your telemetry fits a per-GB budget. Take Datadog instead if many people investigate incidents, because seats at $349 or more each will outrun host pricing.
Grafana Cloud
Grafana Cloud is the managed version of the open-source stack many teams already know: Grafana for dashboards, Mimir or Prometheus for metrics, Loki for logs, Tempo for traces, plus application, Kubernetes and frontend observability built on top. Because it speaks Prometheus and OpenTelemetry, the instrumentation you add is portable: if you leave, your collectors keep working against another backend. That is the main reason Dynatrace users, tired of a proprietary agent, put it on their list.
The free tier is large enough for a real pilot: 10,000 active metric series, 50 GB of logs and 2,232 host hours of application observability a month, which is three hosts around the clock, all kept for 14 days. Pro starts with a $19 monthly platform fee ($228 a year), raises retention to 13 months for metrics and 30 days for logs and traces, then meters each product on its own unit. Metrics cost from $6.50 per 1,000 series beyond the free 10,000. Application Observability starts at $0.025 per host hour, about $18 per host a month, so ten hosts add roughly $180. The trap comes at the top: Enterprise, which brings volume pricing and contract terms, has a minimum commit of $25,000 a year, about $2,083 a month. Our Grafana switch guide lists the reasons some teams move off it, and our Grafana and n8n guide shows how to manage its dashboards, teams and users from a workflow.
- Prometheus and OpenTelemetry compatible from end to end
- Free tier with 10k series, 50 GB of logs and 2,232 host hours
- Self-serve Pro without a sales call
- ✓Instrumentation stays portable if you leave
- ✓Pay-as-you-go Pro with no annual commitment
- ✓13 months of metric retention on Pro
- ✗Many separate meters: series, GB, host hours, sessions
- ✗Enterprise starts at a $25,000 yearly commitment
- ✗Less automatic discovery than Dynatrace, more setup
| Criterion | Grafana Cloud | Dynatrace |
|---|---|---|
| Billing unit | Active series, GB, host hours | Memory-GiB-hours, pods, GiB and sessions |
| Free tier | 10k series, 50 GB logs, 2,232 host hours | 15-day trial |
| Instrumentation | Prometheus and OpenTelemetry | OneAgent, plus OTLP ingest |
Pick Grafana Cloud if portability matters more than automatic discovery and your team can build its own dashboards. Take IBM Instana instead if you want agents that map services for you with almost no setup.
Elastic Observability
Elastic Observability grew out of the ELK stack, and it still shines on logs. Its serverless offer prices on two numbers only, what you ingest and what you keep. Logs Essentials ingests from $0.07 per GB and retains from $0.017 per GB a month; Complete, which adds APM, synthetics and the rest of the observability features, ingests logs and traces from $0.09 per GB. Egress includes 50 GB free, then $0.05 per GB. There are no host fees and no user fees on this model.
Run the numbers on your own volume. At 500 GB of logs a month, ingest costs about $35 on Logs Essentials or $45 on Complete, before retention. Retention is the line that grows: every GB you keep is billed again each month, so a 90-day window means paying for three months of data at once. Elastic backs serverless with a 99.95% monthly uptime SLA. Support is the line to ask about: a Standard subscription gets limited support, while Gold, Platinum and Enterprise add 5%, 10% or 15% of your consumption. If you already run Elasticsearch yourself, the self-managed route avoids the cloud bill but brings back the cluster work; our Elastic alternatives page covers the teams who left for that reason.
- Ingest-and-retain pricing with no host or user fees
- Logs Essentials tier from $0.07 per GB
- 99.95% monthly uptime SLA on serverless
- ✓Search and log analytics at its core, from the ELK stack
- ✓Predictable bill once you know daily volume
- ✓Self-managed option if data must stay in-house
- ✗Retention is billed every month on stored data
- ✗Curated trace analysis, SLOs and synthetics need the Complete tier
- ✗Dashboards and alerting take more setup than Datadog's
| Criterion | Elastic Observability | Dynatrace |
|---|---|---|
| Billing unit | GB ingested and GB retained | Memory-GiB-hours, pods, GiB and sessions |
| Log ingest | From $0.07 per GB | $0.20 per GiB |
| Commitment | Pay as you use on serverless | Minimum annual spend |
Pick Elastic if logs are most of your telemetry and you want a bill that tracks volume alone. Take Datadog instead if APM and infrastructure maps matter more than search.
IBM Instana
IBM Instana is the alternative whose philosophy is closest to Dynatrace: a single agent per host discovers processes, services and dependencies on its own, traces requests end to end and builds the service map without manual tagging. It covers infrastructure, APM, traces and logs in context, with Essentials and Standard bundles, and it can run as SaaS or self-hosted when data must stay inside your walls.
Instana licenses by Managed Virtual Server (MVS): one unit per physical machine, virtual server or worker node, whatever its memory. That is the key difference from Dynatrace, where a 64 GiB host costs eight times an 8 GiB one. The SaaS plan starts at $23.10 per MVS a month, or $277.20 a year, bundled as Essentials (infrastructure) or Standard (full-stack observability), with unlimited users. A standard license has a minimum order of 10 hosts, so about $231 a month at that entry rate. Logs in context are a SaaS add-on from $0.35 per GB. PayPerUse, which needs no time commitment, starts at $0.04 per MVS hour: about $29.20 for a server running all month, $6.10 more than SaaS. Self-Hosted starts at $385.20 per MVS a year, sold as an annual subscription. The 14-day trial gives full access, enough to compare its service map with yours.
- Automatic discovery and tracing, close to OneAgent
- One license unit per server, regardless of memory
- SaaS, pay-per-use or self-hosted with the same features
- ✓Flat per-server price for large-memory hosts
- ✓No-commitment PayPerUse option
- ✓Self-hosting for regulatory and sovereignty needs
- ✗300+ supported technologies, against Datadog's 1,000+ integrations
- ✗PayPerUse runs about $6.10 more per server a month than SaaS
- ✗Licenses sold from a 10-host minimum
| Criterion | IBM Instana | Dynatrace |
|---|---|---|
| Billing unit | Managed virtual server | Memory-GiB-hours |
| 64 GiB host | Same price as any server | 8 times an 8 GiB host |
| Minimum | 10 hosts for a standard license | Minimum annual spend |
Pick Instana if you want Dynatrace-style automation on big hosts without memory-based billing. Take Grafana Cloud instead if you prefer open formats and a free tier over automatic mapping.
SigNoz
SigNoz is an open-source observability platform (its community edition is MIT-licensed, enterprise features sit under a separate license) that puts logs, metrics and traces in one interface, and you can install it yourself with Docker, on Linux or on Kubernetes. For teams that only need APM and logs, not Dynatrace's security and business analytics modules, it can cover what they actually used. The cloud edition removes the hosting work and bills on data only: no host-based pricing, and no special pricing for custom metrics.
Cloud Teams costs $49 a month, and that minimum includes $49 of usage, for example 163 GB of logs or traces. After that, traces and logs cost $0.30 per GB with 15 days of retention, and metrics $0.10 per million samples. Retention changes the price: logs kept for a year cost $1.40 per GB instead of $0.30, so 500 GB a month is $150 at the base rate and $700 on one-year retention. Startups under three years old, with fewer than 30 employees and less than $6 million raised, pay $19 a month for the first 12 months, then the standard $49: a $360 jump over the second year. SSO and audit logs point to the Enterprise plan, which starts at $4,000 a month. Self-hosting is free in license fees, not in time: your team operates the infrastructure, storage, scaling, upgrades and backups.
- Open source, self-hosted with Docker, Linux or Kubernetes
- No host-based pricing on the cloud edition
- One store for logs, metrics and traces
- ✓Entry price of $49 a month with usage included
- ✓Data-only billing, so host count does not matter
- ✓An exit path: the self-hosted edition stays free
- ✗Long retention more than quadruples the per-GB price
- ✗Fewer integrations and less automation than Dynatrace
- ✗Startup discount ends after 12 months
| Criterion | SigNoz | Dynatrace |
|---|---|---|
| Billing unit | GB and metric samples | Memory-GiB-hours, pods, GiB and sessions |
| Entry price | $49 a month with $49 of usage | Minimum annual spend |
| License | MIT community edition, free | Commercial subscription (DPS) |
Pick SigNoz if budget drives the move and your team is comfortable with OpenTelemetry. Take New Relic instead if you want a free tier with more automation and a larger vendor behind it.
Dynatrace alternatives compared
The six alternatives side by side, with the entry price each vendor publishes. Each tool gets five sub-scores out of 5 and its score is their simple average: ease of use, value for money, features and depth, customer support and assistance, available integrations. The "what the bill counts" column is the one most guides skip, and the one that decides whether leaving Dynatrace actually saves money.
| Best for | What the bill counts | Free plan | Team size | Visit | ||||
|---|---|---|---|---|---|---|---|---|
| 1 | Datadog | Best overall Dynatrace alternative | Per host and per product: $15 infra + $31 APM yearly | 4.5/5 | Infrastructure Pro $15 per host a month billed yearly ($18 on demand); APM from $31 per host on top | ✓ | Any size | Visit → |
| 2 | New Relic | Best Dynatrace alternative for small teams with large data | GB above 100 free ($0.40) and full users ($99 to $418.80) | 4.4/5 | Free with 100 GB of data a month; then $0.40 per GB and $99 per additional full user on Standard | ✓ | Up to 5 full users on Standard | Visit → |
| 3 | Grafana Cloud | Best open-standards Dynatrace alternative | Series, GB and host hours; $25,000 floor for Enterprise | 4.3/5 | Free tier; Pro from a $19 platform fee a month, then usage (application observability from $0.025 per host hour) | ✓ | Platform teams | Visit → |
| 4 | Elastic Observability | Best Dynatrace alternative for log-heavy teams | GB ingested ($0.07 to $0.09) and GB retained each month | 4.2/5 | Serverless ingest from $0.07 per GB (Logs Essentials) or $0.09 per GB (Complete), plus retention from $0.017 per GB a month | — | Log-heavy teams | Visit → |
| 5 | IBM Instana | Best Dynatrace alternative for automatic discovery | Managed virtual servers, memory ignored | 4.1/5 | SaaS from $23.10 per managed virtual server a month; PayPerUse from $0.04 per MVS hour | — | Enterprise, big hosts | Visit → |
| 6 | SigNoz | Best open-source Dynatrace alternative | GB and metric samples; retention multiplies the GB price | 4.0/5 | Self-hosted free; Cloud Teams $49 a month including $49 of usage, then $0.30 per GB of logs or traces | ✓ | Startups, OTel teams | Visit → |
Is your tool missing from this page? Email us: hello@hackceleration.com
Which Dynatrace alternative for which case
Infrastructure, APM, logs, RUM and security, 1,000+ integrations.
Your team is small and your data is bigNew Relic100 GB free each month, no per-host fees.
You run Prometheus or OpenTelemetry alreadyGrafana CloudOpen formats, free tier, Pro from $19 a month.
Logs are most of your billElastic ObservabilityIngest from $0.07 per GB, no host fees.
You want OneAgent-style automatic discoveryIBM InstanaOne agent per host, $23.10 per server a month on SaaS.
Budget comes first and you can self-hostSigNozOpen source, or Cloud from $49 a month.
You mainly need uptime and status pagesBetter Stack switch guideWhat a lighter monitoring stack covers, and what it misses.
You also monitor LLM callsLLM observability rankingTools that trace prompts, tokens and costs.
How to choose a Dynatrace alternative
Dynatrace belongs to a category usually called observability platforms, which grew out of APM (application performance monitoring). They all work in a similar way. An agent or an OpenTelemetry collector runs on each host, container or function. It gathers four kinds of signals: metrics (numbers over time, such as CPU or request rate), traces (the path of one request across services), logs (the text lines your code writes) and, on the front end, real user sessions. The agent ships everything to the vendor's backend, which stores it, links signals together and lets you search, chart and alert. What differs is how much the agent discovers on its own, how open its formats are, and above all what the vendor counts to bill you.
Learn the units before you compare prices. A host is one server, VM or Kubernetes node, the meter of Datadog's infrastructure and APM. A host hour is one host monitored for one hour, so a server running all month is about 730 host hours; Grafana Cloud and Dynatrace's Infrastructure Monitoring use it. A memory-GiB-hour multiplies those hours by the host's memory, which is how Dynatrace bills Full-Stack. A GiB or GB ingested is data volume sent in, the meter of New Relic, Elastic and SigNoz; GB retained is what you keep, billed again each month. An active series is one metric with one set of labels, Grafana's metrics meter. An MVS is one managed server on Instana. A full user is a seat on New Relic.
Then demand four answers from each vendor. What happens when you go over: same rate, a penalty rate, or a pause? Is the price yearly or monthly, and is it a first-year rate? Which signals sit on the entry plan, and which need a higher tier? Can you export your dashboards and keep your instrumentation if you leave? Several names from other lists stay out of the ranking. Splunk AppDynamics appears in two guides, but it brings back the annual contract: Premium starts at $33 per host a month billed annually, and Splunk's observability editions start at 30 hosts. Site24x7, also in two guides, is built for smaller estates: its Professional plan, $42 a month billed annually, covers 5 servers and a single application. Splunk Observability Cloud, named by three of the top results, lists Essentials at $95 per host a month and starts at 30 hosts, about $2,850 a month. Sentry is built around your code (errors, traces, logs and session replays, Team from $26 a month billed annually) and pairs well with any tool here; our Sentry switch guide explains its limits. Honeycomb, from $150 a month for 50 million events, suits teams that live in traces. Splunk's broader platform is covered in our Splunk switch guide.
Ask how AI investigations are billed, because root-cause analysis, the selling point of Dynatrace's Davis engine, is now metered almost everywhere. Dynatrace bills Agentic Operations at $0.10 per AI unit, and an autonomous multi-step investigation averages 30 units, about $3; the same tasks inside Dynatrace Assist are included for a limited time. Datadog sells AI credits for its Bits AI features from $500 per 500 credits a month billed annually, or $1.30 per credit on demand. Grafana Cloud includes 3 active AI users a month on its free tier, then charges $20 per active AI user. New Relic bills its AI features as Advanced Compute units. If you would rather query your telemetry from your own assistant, our guides to connecting Claude to Datadog or to Grafana show what each connector can read and change.
A startup with a few services
A platform team on Kubernetes
An enterprise with large hosts
A team drowning in logs
- Find out what the bill counts: hosts, host hours, memory, GB or users
- Ask what happens above the commitment: same rate, premium or pause
- Compare yearly and monthly prices on the same plan
- Check which signals the entry plan includes
- Price your largest hosts, not an average one
- Ask how AI root-cause investigations are metered
- Make sure dashboards and instrumentation can leave with you
Dynatrace alternatives · FAQ
What happens if you use more than your Dynatrace commitment?
You keep using the platform. Usage beyond the yearly commitment is billed monthly on demand at the same rates as your prepaid consumption, with no penalty-style overage. That makes a spike costly but never a cut-off, so set budget alerts on each capability.
How much can you save by switching from Dynatrace?
It depends on host memory more than on host count. On 8 GiB hosts, Dynatrace Full-Stack lists $58 a month and Datadog Infrastructure Pro plus APM $46 billed yearly, a $12 gap per host. On 32 GiB hosts, Dynatrace reaches about $232 against the same $46 at Datadog. Compare like for like: Full-Stack includes 67 GiB of trace volume and Kubernetes platform monitoring, while Datadog bills logs on their own meters.
Does Dynatrace support OpenTelemetry?
Yes. Dynatrace ingests metrics and traces over OTLP, and Container Application Monitoring, at $5 a month per container, covers workloads instrumented with OpenTelemetry or with Dynatrace. That helps when you leave: services you instrument with OpenTelemetry today can later point at Grafana Cloud, SigNoz or Datadog by changing the exporter, while OneAgent's automatic discovery stays behind.
Can you run another APM tool alongside Dynatrace during a migration?
Yes, and it lowers the risk of the switch. An OpenTelemetry Collector can send the same telemetry to one or more open source or commercial backends, so you can feed Dynatrace and the new tool in parallel and compare their alerts before you cut over. Grafana also publishes a Dynatrace data source plugin, listed as a Grafana Enterprise plugin, that queries Dynatrace metrics from inside Grafana while both run.
Can you migrate from Dynatrace without losing historical data?
Partly, and it depends on what Dynatrace kept. Metrics come with 15 months of retention at 1-minute granularity, traces with 10 days unless you paid to extend them, and logs with the window you set, from 1 day to 10 years. Without extended retention, traces older than ten days are already gone, so start the new tool before the contract ends and let it build its own history.
Does Dynatrace charge per user?
No. A Dynatrace subscription includes unlimited seats, with no per-seat fees and no user tiers, and IBM Instana SaaS also includes unlimited users. New Relic works the other way: 30 full users on Pro cost $10,470 a month on an annual commitment before any data, a line Dynatrace does not bill at all.
What are Dynatrace host units, DDUs and DEM units?
They are the units of Dynatrace's previous license model, which you still meet on older contracts and quotes. Host units measured each OneAgent-monitored host, Davis data units (DDUs) licensed capabilities such as custom metrics and traces, DEM units covered digital experience monitoring and ASUs application security. The current Dynatrace Platform Subscription bills a dollar rate card instead, so convert an old quote before comparing it with any tool here.
Does Dynatrace charge for each real user session?
Yes. Real User Monitoring is billed at $0.00225 per session, which is $2.25 per 1,000 sessions, with 35 days of data retention. Sessions with replay capture cost twice as much, $0.0045 each, so enable Session Replay only on the pages where you need it.
Does Dynatrace include application security?
As paid add-ons on the same subscription. Runtime Vulnerability Analytics and Runtime Application Protection each cost $13 a month per 8 GiB host, billed per memory-GiB-hour, and the second blocks exploitation attempts against top OWASP risks. Security Posture Management costs $5 a month per host.