Fazlur Rahman
Last verified: October 7, 2026.
Dynatrace's pricing in 2026 is the most metered of any major APM vendor. The Dynatrace Platform Subscription (DPS), introduced in August 2024 and now the only commercial path, charges every consumption dimension on its own: full-stack monitoring bills per memory-GiB-hour, infrastructure bills per host-hour, Kubernetes bills per pod-hour, logs bill per GiB ingest plus per GiB-day retention, traces bill per GiB plus per GiB scanned, RUM bills per session, security bills per memory-GiB-hour. The math that emerges is the real cost of Dynatrace, and it is meaningfully different from the headline price of $29/host for Infrastructure or $58/8 GiB for Full-Stack.
This walkthrough reads Dynatrace the way a finance team encounters the first invoice: the rate card line by line, what each capability actually bills, what a representative production environment costs end-to-end, the price tier that often gets overlooked (Foundation & Discovery), the DPS commitment math, and the comparison against Datadog and New Relic that determines whether Dynatrace is the right vendor for the workload.
How the 2026 Rate Card Works
DPS replaced the older per-host + per-GiB split that defined Dynatrace's pricing through 2023. Under DPS, customers make one annual commitment at the platform level — not per capability — and consume from that commitment at the rate-card prices:
| Capability | Unit | Rate |
|---|---|---|
| Foundation & Discovery | per host-hour | $0.01 |
| Infrastructure Monitoring | per host-hour | $0.04 |
| Full-Stack Monitoring | per memory-GiB-hour | $0.01 |
| Database Monitoring | per database-instance-hour | $0.11 |
| Mainframe Monitoring | per MSU-hour | $0.10 |
| Kubernetes Platform Monitoring | per pod-hour | $0.002 |
| Code Monitoring | per container-hour | $0.005 |
| Log Ingest & Process | per GiB | $0.20 |
| Log Retain (with queries) | per GiB-day | $0.02 |
| Log Retain (storage-only) | per GiB-day | $0.0007 |
| Log Query | per GiB-scanned | $0.0035 |
| Trace Ingest & Process | per GiB | $0.20 |
| Trace Retain | per GiB-day | $0.0007 |
| Trace Query | per GiB-scanned | $0.0035 |
| Custom Metrics Ingest | per 100k datapoints | $0.15 |
| Runtime Vulnerability Analytics | per memory-GiB-hour | $0.00225 |
| Runtime Application Protection | per memory-GiB-hour | $0.00225 |
| Real User Monitoring | per session | $0.00225 |
| RUM Session Replay | per session | $0.0045 |
| Browser / Clickpath Synthetic | per synthetic action | $0.0045 |
| Automation Workflow | per workflow-hour | $0.03 |
| Data Egress | per GiB | $0.15 |
The hourly pricing is industry-unique: Dynatrace bills in fractions of an hour rather than per day, per month, or per high-water-mark. A host that runs for ten hours bills for ten hours. The approach rewards elastic workloads and punishes always-on workloads that grow.
What Full-Stack Monitoring Actually Costs
The Full-Stack rate of $0.01 per memory-GiB-hour is the single largest line on most production bills. The math:
One 8 GiB host running 24/7: 8 GiB × 730 hours = 5,840 GiB-hours. At $0.01/GiB-hour = $58.40/month. The headline $58/month per 8 GiB host is exact — Dynatrace publishes it as the canonical reference price.
One 16 GiB host running 24/7: 16 GiB × 730 = 11,680 GiB-hours × $0.01 = $116.80/month. Doubling RAM doubles the bill.
The 4 GiB minimum is the floor. A 2 GiB host bills as 4 GiB; the bill is met if the host has 4 GiB or less. The 0.25 GiB rounding means a 4.1 GiB host bills as 4.25 GiB.
Production reality: most modern application servers run at 16 GiB or higher, and Kubernetes node sizes of 32 GiB or 64 GiB are typical in containerized environments. A 100-host production fleet at 32 GiB average RAM = 3,200 GiB × 730 hours × $0.01 = $23,360/ month for Full-Stack alone.
The Foundation & Discovery Tier
Foundation & Discovery is the pricing tier most evaluations miss. It is positioned below Infrastructure Monitoring and bundles:
| Capability | Included |
|---|---|
| Basic host health indicators | Yes |
| Host and process topology detection | Yes |
| Filesystem monitoring (disk usage) | Yes |
| OS services monitoring | Yes |
| Collection and correlation of logs in context | Yes |
| Application Performance Monitoring (APM) | No |
| Root-cause analysis | No |
| Code-level profiling | No |
Foundation & Discovery is $0.01 per host-hour = $7/month per host (dynatrace.com/pricing). It is the right tier for OS-level visibility on hosts that do not run Dynatrace-monitored applications — dev/staging boxes, database servers with separate Database Monitoring, and unmonitored coverage that teams want to see but do not need full APM on. The Discovery & Coverage app that ships with Foundation lets teams find unmonitored hosts automatically and assess whether they need upgrading.
What Logs Cost Separately
Logs are not included in the host allowance. A 100-host production environment with 5 TB of monthly log ingest pays $1,000/month just for ingest (5,000 GiB × $0.20/GiB), plus retention. Retain-with-included-queries at $0.02/GiB-day for 30 days = $3,000/month. Total log cost on a typical production fleet: $4,000/month on top of Full-Stack.
The math matters because most evaluations model only the Full-Stack line and miss the logs line. The OpenObserve Dynatrace comparison calculator estimates a 5 TB log volume at $4,000/month log spend on top of $11,680 Full-Stack — total $17,127/month before traces, RUM, and security.
Log retention duration is the variable. The "retain with included queries" tier caps included queries at 10–35 days; longer retention runs further. Teams that need 90-day log retention for compliance pay roughly $9,000/month on 5 TB ingest — three times the 30-day retention cost.
Real Production Bill: 100-Host 16-GiB Environment
A representative end-to-end bill for a 100-host production environment at 16 GiB average RAM, 5 TB monthly logs, 2.5 TB traces, 85 GB metrics, 15 GB RUM, and 15 active workflows:
| Capability | Monthly cost |
|---|---|
| Full-Stack (100 hosts × 16 GiB × 730 h × $0.01) | $11,680 |
| Infrastructure-Only Hosts (additional 50 hosts at $29/mo each) | $0 (covered by Full-Stack) |
| Kubernetes pods (non-FSM, 100 pods × $0.002 × 730) | $146 |
| Logs (5 TB ingest + 30d retain-with-queries) | $4,000 |
| Traces (2.5 TB ingest + 30d retain) | $502 |
| Metrics (85 GB / month) | $12 |
| RUM (15 GB sessions) | $1,687 |
| Active workflows (15 × $0.03 × 730) | $329 |
| Data queries (1 TB scanned × $0.0035) | $3.50 |
| List price subtotal | $18,360 |
| Enterprise support (+10%) | $1,836 |
| Total monthly list price | $20,196 |
Enterprise customers negotiate volume discounts on the DPS commitment; list price rarely ships. A typical 30% enterprise discount brings $20,196/month to roughly $14,137/month — about $169,000/year for a 100-host Full-Stack production environment.
The Comparison That Decides the Vendor
Dynatrace competes directly with Datadog and New Relic in the APM market. The pricing model differences are stark:
| Vendor | Primary metric | 100-host list price (annual) | Model |
|---|---|---|---|
| Dynatrace DPS | memory-GiB-hour | $170,000–$240,000 | Per-GiB-hour, all signals metered separately |
| Datadog | host + indexed logs | $200,000–$280,000 | Per-host + per-GB logs indexed + per-million spans |
| New Relic | user + data ingest | $150,000–$220,000 | Per-user + per-GB data ingest |
| Splunk Observability | hosts + custom metrics | $250,000+ | Per-host + per-100 custom metrics |
Dynatrace wins on AI-driven root-cause analysis (the Davis AI engine is the long-running differentiator), on auto-instrumentation without code changes for most stacks, and on Kubernetes coverage depth. Datadog wins on the product surface area (20+ products vs Dynatrace's narrower set), on per-host simplicity, and on log analysis depth. New Relic wins on price for organizations under 100 hosts and on the free tier (100 GB/month free, 1 full user). Splunk wins on enterprise compliance posture and on the OT/SRE handoff.
When Dynatrace Is the Right Pick
Dynatrace is the right APM vendor when you have 50+ hosts running complex applications, need automated root-cause analysis without manual instrumentation, are running Kubernetes at scale, and want the AI-driven correlation that Davis provides. The DPS commitment math rewards annual commitments over monthly billing; customers who commit annually and stay under the commitment save 10–25% on list.
Consider alternatives when you have under 30 hosts (the per-host math favors Datadog and New Relic), need heavy log analysis at scale (Splunk and Datadog have deeper log products), or operate under 50 GB/day ingest (New Relic's free tier covers this). For DevOps teams already running Kubernetes at 100+ nodes and needing end-to-end visibility, Dynatrace is the strongest product on the market and the 2026 DPS model is the cleanest pricing it has published.
What to Budget For
Three budget anchors for Dynatrace sizing:
Small (10 hosts, 8 GiB average, 1 TB logs): roughly $600/month Full-Stack + $200/month logs = $800/month list. Enterprise discount brings to roughly $560/month.
Mid (50 hosts, 16 GiB average, 3 TB logs): roughly $5,840/month Full-Stack + $1,800/month logs = $7,640/month list. Discounted: $5,348/month.
Large (200 hosts, 32 GiB average, 10 TB logs): roughly $46,720/month Full-Stack + $6,000/month logs = $52,720/month list. Discounted: $36,904/month.
Add $1,000–$3,000/month for RUM, security, and Synthetic Monitoring depending on application surface. The "headline $29/host" Infrastructure Monitoring is rarely what production customers pay; the Full-Stack rate-card math plus logs is the realistic budget anchor.
Read next
New Relic Pricing 2026: Per-Host, Per-GB Hybrid Model is the per-user-plus-ingest alternative for organizations under 100 hosts, and Datadog Pricing 2026: Per-Host vs Usage Pricing Tiers is the per-host-plus-usage comparison that defines the mid-market APM market alongside Dynatrace.
