Quick Summary: Datadog uses a modular, usage-based pricing model where costs vary by product: Infrastructure Monitoring starts at $15-$23/host/month, APM costs $31/host/month, and Log Indexing starts at $1.70/million events indexed. According to official pricing documentation, billing is calculated using high watermark plans that measure the 99th percentile of monthly usage to reduce the impact of spikes.
Datadog bills arrive with little warning. Teams add a few extra dashboards, spin up new services, or increase logging verbosity, and suddenly finance is asking pointed questions.
The platform offers powerful observability capabilities, but its pricing model operates differently than traditional monitoring tools. Each product module bills separately, and usage compounds quickly across distributed systems.
This breakdown covers what Datadog actually costs when running production workloads. Real numbers, not marketing estimates.
How Datadog’s Pricing Model Works
Datadog operates on a modular, usage-based pricing structure. Organizations pay separately for each capability they enable, whether that’s infrastructure monitoring, application performance tracking, or log management.
According to the official pricing documentation, Datadog offers three billing frequencies:
- Annual billing: Lowest per-unit costs with volume commitments
- Monthly billing: Mid-tier pricing with monthly commitments
- On-demand billing: Highest per-unit costs, no commitments
The pricing differs substantially between these tiers. Infrastructure Monitoring Pro costs $15 per host monthly when billed annually, but jumps to $18 for month-to-month or on-demand billing.
High Watermark Billing Explained
Most Datadog products use high watermark plan (HWMP) billing. Here’s how it works:
Datadog records usage metrics every hour throughout the month. At month’s end, they calculate the 99th percentile of that usage—meaning they exclude the top 1% of usage spikes. This maximum value becomes the billable amount.
The official documentation states: “The billable count of hosts is calculated at the end of the month using the maximum count (high-water mark) of the lower 99 percent of usage for those hours. Datadog excludes the top 1 percent to reduce the impact of spikes in usage on your bill.”
This model protects against temporary spikes but means organizations pay for peak capacity, not average usage.
Infrastructure Monitoring Pricing
Infrastructure Monitoring represents the foundation of most Datadog deployments. It tracks hosts, containers, and system-level metrics across an organization’s infrastructure.
According to the official pricing page, Infrastructure Monitoring offers two tiers:
| Plan | Annual (per host/month) | Monthly (per host/month) | On-Demand (per host/month) |
|---|---|---|---|
| Pro | $15 | $18 | $18 |
| Enterprise | $23 | $27 | $27 |
A host is defined as a physical or virtual operating system instance. Datadog records unique hosts hourly, counting each distinct machine running the Datadog Agent.
Container Monitoring Costs
Container monitoring adds another pricing dimension. According to official documentation, containers cost $1 per container per month, or $0.002 per container hour.
The official pricing states organizations receive an allotment of 10 containers per host per hour. This matters for Kubernetes environments where pod counts fluctuate—those containers beyond the included allotment trigger additional charges.
Custom Metrics Pricing
Custom metrics represent one of the most common cost escalation points. The official billing documentation explains the structure:
- Pro accounts: 100 indexed custom metrics per host included
- Enterprise accounts: 200 indexed custom metrics per host included
- Overage charges: $5 per 100 custom metrics per month
The challenge? Modern applications generate custom metrics rapidly. Prometheus exporters, application instrumentation, and third-party integrations all contribute. A single microservice might expose many metrics across multiple label combinations.
Application Performance Monitoring (APM) Costs
APM tracks distributed traces, allowing teams to monitor request flows through microservices architectures. The pricing differs significantly from infrastructure monitoring.
According to the official APM billing documentation, APM costs $31 per APM host per month. Each APM host includes:
- 1 million Indexed Spans per month
- 150 GB of Ingested Spans per month
APM hosts are calculated separately from infrastructure hosts. A machine running both Infrastructure and APM monitoring triggers charges for both products.
Span Ingestion vs. Indexing
Datadog separates span costs into two categories:
- Ingested Spans represent the raw volume of traces sent to Datadog before any filtering. Organizations can ingest up to 150 GB per APM host monthly within their base cost.
- Indexed Spans are the traces retained for search and analysis. The base APM cost includes 1 million indexed spans per host monthly. Additional indexed spans cost extra, though official documentation doesn’t publicly list the overage rate in the standard pricing page.
This dual structure matters because high-throughput applications generate millions of traces. Not all require indexing—many teams sample aggressively to control costs.
Log Management Pricing
Log Management operates on a two-tier pricing model that bills separately for ingestion and indexing.
According to the official pricing page, Log Management charges:
- Ingestion: Starting at $0.10 per GB of ingested or scanned data
- Log Indexing (15-day retention): $1.70 per million events (annual)
- Log Indexing (on-demand): $2.00 per million events
The official documentation clarifies that ingestion costs apply “per GB of uncompressed data ingested for processing, or compressed data scanned for rehydration.”
Understanding Log Retention Tiers
Datadog offers multiple retention options beyond the standard 15-day tier. Organizations can configure retention filters to route different log categories to different tiers, though exact pricing for extended retention isn’t detailed in the public pricing documentation.
The architecture allows teams to:
- Ingest all logs for live tailing and metrics extraction
- Index only critical logs for search and analysis
- Archive everything to self-hosted storage for compliance
This flexibility helps control costs, but requires thoughtful configuration. Teams often discover they’re indexing more than necessary, particularly verbose debug logs that provide little search value.
| Log Type | Typical Volume | Recommended Strategy |
|---|---|---|
| Application errors | Low | Index all (15+ days) |
| Access logs | Very High | Ingest only, archive |
| Debug logs | High | Sample heavily |
| Security events | Medium | Index all, extend retention |
Additional Product Pricing
Beyond the core three products, Datadog offers numerous specialized monitoring capabilities, each with separate pricing.
Database Monitoring
Database Monitoring tracks query performance, execution plans, and database-level metrics. Pricing varies but typically bundles with Infrastructure or APM tiers for specific database types.
Real User Monitoring (RUM)
RUM tracks frontend performance and user sessions. According to official documentation, Datadog charges based on session volume, with separate billing for standard RUM and RUM Replay (session recording).
Synthetic Monitoring
Synthetic tests simulate user interactions to monitor availability and performance. The official pricing indicates costs start at approximately $12 per 1,000 browser test runs per month, with API tests priced separately.
Security Products
Datadog’s security suite includes Cloud Security Posture Management (CSPM), Cloud Workload Security (CWS), and Application Security Management (ASM). These products bundle into DevSecOps packages:
- DevSecOps Pro: $22 per host per month (annual)
- DevSecOps Enterprise: $34 per host per month (annual)
Error Tracking
Error Tracking aggregates application errors across services. According to the official pricing page, it costs:
| Volume Tier | Annual (per 1k errors) | Monthly (per 1k errors) |
|---|---|---|
| 50k – 100k errors | $0.25 | $0.30 |
Workflow Automation
Workflow Automation enables automated responses to monitoring events. The official billing documentation states that CSM Pro includes 5 workflow executions per host monthly, while CSM Enterprise includes 20 executions per host monthly. Additional executions trigger overage charges.
What Drives Costs Higher
Datadog bills grow through predictable patterns. Understanding these helps teams anticipate expenses.
Microservices Multiplication
Each service that needs monitoring adds hosts, containers, custom metrics, and APM traces. A monolith might run on 10 hosts. The same application split into 30 microservices could require 50+ hosts when accounting for redundancy.
The container allotment (10 per host per hour) helps, but large Kubernetes clusters quickly exceed this. A cluster running 500 pods across 20 nodes generates container monitoring charges for 480 containers beyond the included allotment.
Observability Without Limits
Verbose logging represents another common escalation point. Applications configured to log at DEBUG level in production generate massive volumes. At $0.10 per GB ingested, a service producing 1 TB of logs monthly costs approximately $100 in ingestion charges at the standard rate—before indexing charges.
Metric Cardinality Explosions
Custom metrics multiply through tag combinations. A metric with five tag keys, each having ten possible values, generates up to 100,000 unique metric combinations. Datadog bills per unique metric time series.
Common culprits include user IDs, request IDs, or container hashes used as metric tags. These high-cardinality labels create exponential growth in billable metrics.
Development and Testing Environments
Organizations often instrument non-production environments identically to production. Three environments (dev, staging, production) triple the host count. Datadog charges the same rate regardless of environment.
Hidden Cost Factors
Beyond the obvious per-unit pricing, several factors impact total cost of ownership.
Data Retention Requirements
Compliance mandates often require extended log retention. While Datadog supports archiving to self-hosted storage, rehydrating those logs for analysis triggers scanning charges at the ingestion rate.
On-Demand Overages
Organizations on committed plans that exceed their contracted volumes pay on-demand rates for overages. These rates can be 20-50% higher than committed pricing.
Alert Fatigue and Noise
Comprehensive monitoring generates alerts. Managing alert noise requires additional integrations, escalation policies, and potentially separate incident management tooling—though Datadog does offer Incident Management with its own pricing structure.
Cost Management Strategies
Organizations can control Datadog spending through several approaches.
Tag-Based Resource Allocation
Proper tagging enables cost visibility. Tagging hosts and services by team, environment, and cost center allows finance teams to allocate expenses accurately. Datadog’s Cost Details page supports filtering by these tags.
Metrics Without Limits Configuration
The official custom metrics documentation describes Metrics without Limits as a way to reduce indexed metric volume. Teams configure which tag combinations to retain, discarding high-cardinality dimensions that provide limited value.
A metric tracking HTTP request duration by endpoint, status code, region, and customer ID might only need retention for endpoint and status code. Dropping customer ID from indexing reduces cardinality exponentially while preserving operational insights.
Log Sampling and Filtering
Not every log needs indexing. Teams can configure exclusion filters to:
- Drop health check logs entirely
- Sample verbose debug logs at 1%
- Index only logs containing error keywords
- Route low-priority logs to cheaper retention tiers
The official Log Management documentation details these filtering capabilities, noting that exclusion filters apply before indexing charges.
APM Trace Sampling
High-throughput services generate millions of traces. Aggressive sampling captures sufficient data for performance analysis while controlling span ingestion costs. Many teams find that sampling at 1-5% of total traffic provides sufficient data for baseline performance monitoring.
Environment Segregation
Organizations can separate development environments into different Datadog accounts with lower-tier plans or reduced monitoring scope. Production runs the full observability stack while development uses minimal infrastructure monitoring.
Lower Datadog Monitoring Expenses with Available Credits
Datadog pricing increases with usage – hosts, logs, metrics, and integrations all add to the bill as your infrastructure grows. Once systems move into production, monitoring can quickly become a significant recurring cost, especially for startups handling real traffic.
Get AI Perks aggregates credits and discounts for more than 200 AI, SaaS, and developer tools, including Datadog offers such as up to $100,000 in credits with one year free and around $5,000 via AWS programs with trial and discounts. The platform shows conditions, approval likelihood, and how to apply, helping founders quickly identify which programs are worth pursuing.
Before committing to higher monitoring spend, review the available Datadog perks on Get AI Perks and secure any credits you can access.
When Datadog Makes Financial Sense
Despite the costs, Datadog provides value in specific scenarios.
Organizations with complex distributed systems benefit from unified observability. Managing separate tools for infrastructure, APM, logs, and RUM creates integration overhead and correlation challenges. Datadog’s single-pane approach reduces operational complexity.
Teams that need rapid incident response value the platform’s correlation capabilities. During outages, jumping between disconnected monitoring tools wastes critical time. Datadog connects metrics, traces, and logs within single views.
Companies with limited observability expertise benefit from Datadog’s integrations. The platform provides pre-built dashboards and monitors for 650+ technologies. Setting up equivalent observability with open-source tools requires substantial engineering time.
Frequently Asked Questions
How does Datadog calculate monthly host counts?
According to official billing documentation, Datadog records unique hosts every hour throughout the month. At month’s end, they calculate the 99th percentile of hourly measurements and bill based on that high watermark. This means the top 1% of usage spikes are excluded from billing, reducing the impact of temporary capacity increases.
What’s the difference between ingested and indexed spans in APM?
Ingested spans represent all traces sent to Datadog before filtering. The official APM billing documentation states each APM host includes 150 GB of ingested spans monthly. Indexed spans are traces retained for search and analysis, with 1 million indexed spans included per APM host monthly. Organizations can ingest all traces for real-time analysis while indexing only a subset for historical search.
Do custom metrics charges apply to standard metrics?
No. According to the official custom metrics documentation, integration metrics from Datadog’s 650+ integrations don’t count toward custom metric limits. Only metrics sent via DogStatsD, the Datadog API, or custom checks count as custom metrics. Pro accounts include 100 indexed custom metrics per host, while Enterprise accounts include 200.
How much do containers cost beyond the included allotment?
The official pricing page lists container monitoring at $1 per container per month or $0.002 per container hour. Each infrastructure host includes an allotment of 10 containers per host per hour. Containers beyond this allotment trigger additional charges. In a Kubernetes environment with high pod density, this can add substantial costs.
Can log ingestion and indexing be configured separately?
Yes. The official Log Management documentation explains that organizations can ingest all logs for live tailing and metrics extraction while indexing only specific log categories. Exclusion filters applied during ingestion prevent logs from being indexed, avoiding the $1.70 per million events indexing charge while still allowing logs to flow through processing pipelines.
What’s included in the DevSecOps pricing bundles?
DevSecOps bundles combine infrastructure monitoring with security products. According to official pricing, DevSecOps Pro costs $22 per host monthly (annual billing) and includes Cloud Security Posture Management and Cloud Workload Security. DevSecOps Enterprise costs $34 per host monthly and adds Application Security Management and additional workflow automation executions.
How does Datadog bill for serverless functions?
The official billing documentation states Datadog bills serverless based on the average number of functions per hour across the month. Each hour, Datadog records functions that are executed at least once. At month’s end, charges are calculated from the average hourly function count. This differs from host-based billing and can result in lower costs for infrequently-invoked functions.
Making the Datadog Decision
Datadog pricing reflects its position as an enterprise observability platform. The costs are substantial, but predictable once teams understand the billing model.
Organizations should calculate expected costs based on actual infrastructure footprints before committing. A 100-host deployment with APM and log management could reach $5,000-$8,000 monthly depending on log volume and configuration. At large scale with comprehensive monitoring across all products, bills can exceed six figures.
The platform delivers value through reduced operational overhead and faster incident response. But that value must justify the cost. Teams should evaluate whether they need the full observability stack or if targeted monitoring for specific components suffices.
For organizations committed to Datadog, cost management becomes an ongoing practice. Regular reviews of custom metric cardinality, log indexing patterns, and host optimization prevent surprise bills. The platform provides usage monitoring tools—teams just need to use them.
Ready to get started? Visit the official Datadog pricing page to configure an estimate based on your infrastructure, or start with the free tier to evaluate capabilities before committing to paid plans.
