Datadog vs New Relic vs Grafana Cloud Pricing 2026: Full Cost Breakdown
If you're choosing an observability platform in 2026, the pricing models have never been more different. Datadog charges per host per module. New Relic charges per GB ingested. Grafana Cloud offers a free tier that's 50x more generous than Datadog's. Here's exactly what each platform costs at real team sizes.
Key insight: Grafana Cloud's free tier includes 10,000 metrics, 50GB logs, and 50GB traces. Datadog's free tier? 5 hosts and 1-day retention. For startups and small teams, Grafana Cloud is the clear winner. For enterprises that need turnkey APM, Datadog's per-host model is simpler to predict — but 5–10x more expensive.
Datadog Pricing 2026: The Module Trap
Datadog's pricing model remains the most complex — and most expensive — of the three. You pay separately for almost every product: Infrastructure, APM, Log Management, Real User Monitoring, Synthetics, Security, and more. A typical engineering team uses 4–6 products simultaneously, and each one adds to the bill.
Datadog Infrastructure Monitoring (per host/month)
| Plan | Monthly Billing | Annual Billing | What's Included |
|---|---|---|---|
| Free | $0 (5 hosts, 1-day retention) | $0 | Infrastructure metrics only, no APM |
| Pro | $15/host/mo | $12/host/mo | 15-month retention, dashboards, alerts |
| Enterprise | $23/host/mo | $18/host/mo | Everything + security, compliance |
Datadog APM + Distributed Tracing (additional)
| Plan | Cost per Host/Month | Included Spans |
|---|---|---|
| APM + Profiling | $40/host/mo | 1M indexed spans/month |
| Additional indexed spans | $0.0015 per 1K spans | — |
| Ingested spans | $0.10 per GB ingested | — |
Datadog Log Management (additional)
| Action | Cost |
|---|---|
| Log ingestion | $0.10 per GB (first 15 days) |
| Log indexing (15-day retention) | $1.80 per million log events |
| Log indexing (30-day retention) | $2.50 per million log events |
| Log archive (S3) | Your S3 costs + $0.05/GB rehydration |
The Datadog billing shock: A team ingesting 100 GB/day in logs pays $10/day = $300/month just on log ingestion — before indexing. 100 GB/day × 30 days = 3,000 GB × $0.10 = $300/mo. Index those logs for 30 days and you add 1.8M log events × $2.50/M = $4,500/month. Log management alone can easily hit $5,000+/month for a busy system. The same volume on Grafana Cloud's Pro plan would cost roughly $900/month.
Other Datadog Products (per host or per usage)
| Product | Cost |
|---|---|
| Real User Monitoring (RUM) | $1.50 per 1K sessions |
| Synthetics (browser tests) | $7.20 per 10K test runs |
| Database Monitoring | $70/database host/month |
| Cloud Security (CSPM) | $15.5/host/mo (annual) |
| Incident Management | $10/user/mo |
| Workflow Automation | $0.20 per workflow execution |
New Relic Pricing 2026: Per-GB Model
New Relic restructured to a per-GB data ingestion model in 2021. The core model: you pay for data in, not servers monitored. This makes costs more predictable for teams with many hosts but moderate log volume.
New Relic Core Model
| Component | Free | Standard | Pro | Enterprise |
|---|---|---|---|---|
| Monthly data ingest | 100 GB free | $0.30/GB after 100 GB | $0.25/GB after 100 GB | Volume discounts |
| Users (full platform) | 1 free | $0.10–$0.35/hr (CCU) | $0.10–$0.35/hr | Volume pricing |
| Core users (limited) | Unlimited | Free | Free | Free |
| Data retention | 8 days | 30 days | 90 days | 395 days |
| Features | All (limited scale) | Full | Full + SLA | Enterprise features |
New Relic User Types Explained
New Relic has three user types with very different costs:
- Core users: Free. Can view dashboards, alerts, and use basic features. Cannot use APM tracing details or curated experiences.
- Full platform users: $0.10–$0.35/hr (concurrent user model). Most developers interacting with the platform need this tier.
- Basic users: Free. Read-only access to everything. Useful for stakeholders who just view dashboards.
New Relic billing complexity: New Relic charges per concurrent user-hour, not per seat. If your team has 5 engineers who each use New Relic for 8 hours/day, that's 40 concurrent user-hours/day × ~$0.25/hr = $10/day = $300/month just in user costs. Usage varies wildly depending on oncall rotations and incident volume.
Grafana Cloud Pricing 2026: Open Source Meets Cloud
Grafana Cloud is built on the open-source Grafana, Prometheus, Loki, and Tempo stack. The pricing model is usage-based with a generous free tier that makes it the cheapest option for most small-to-medium teams. The key differentiator: you can self-host the entire stack for free if you want to manage it yourself.
Grafana Cloud Plans
| Component | Free | Pro ($29/user/mo) | Advanced (Custom) |
|---|---|---|---|
| Metrics (Prometheus) | 10,000 series free | $8 per 1K series/mo | Volume pricing |
| Logs (Loki) | 50 GB/month free | $0.50 per GB | Volume pricing |
| Traces (Tempo) | 50 GB/month free | $0.50 per GB | Volume pricing |
| Users | Unlimited viewers, 3 editor seats | $29/user/mo | Custom |
| Retention | 14 days | 13 months | Custom |
| Support | Community forums | Email support | 24/7 SLA |
Why Grafana Cloud's Free Tier Matters
For context on how generous Grafana Cloud's free tier is:
- 10,000 metrics series covers a small-to-medium deployment. Datadog's free tier is 5 hosts — most teams exceed that on day one.
- 50 GB logs/month is enough for most startups. New Relic offers 100 GB free, but charges per user-hour on top.
- 50 GB traces/month is effectively unlimited APM for small teams. Datadog charges $40/host/month for APM alone.
- Unlimited viewer users means your entire team can see dashboards without paying per-seat costs.
Self-hosting option: If you're willing to manage your own infrastructure, you can run the entire Grafana stack (Grafana, Prometheus, Loki, Tempo) on your own servers for $0. This is the ultimate cost optimization — but requires DevOps expertise to set up and maintain. Most teams start with Grafana Cloud and migrate to self-hosted later if costs become an issue.
Grafana Cloud Cost Examples
| Usage Level | Metrics | Logs | Traces | Monthly Cost |
|---|---|---|---|---|
| Startup (5 engineers) | 8,000 series | 30 GB | 20 GB | $0 (free tier) |
| Growing (20 engineers) | 25,000 series | 150 GB | 100 GB | $345/mo (Pro) |
| Scale (50 engineers) | 80,000 series | 500 GB | 300 GB | $1,365/mo (Pro) |
Head-to-Head: 5-Engineer Team
A 5-engineer startup with 10 hosts, 50 GB/month logs, and moderate APM usage.
Datadog — 5 Engineers, 10 Hosts
Infrastructure (10 hosts × $15)$150/mo
APM (10 hosts × $40)$400/mo
Log Management (50 GB × $0.10)$5/mo
Log Indexing (50 GB ≈ 5M events × $1.80)$9/mo
Total$564/mo ($6,768/yr)
New Relic — 5 Engineers, 50 GB Logs
Data ingest (50 GB, within free tier)$0/mo
5 full users × 8 hrs/day × $0.25/hr × 22 days$220/mo
Total$220/mo ($2,640/yr)
Grafana Cloud — 5 Engineers, 50 GB Logs
Metrics (8,000 series, within free tier)$0/mo
Logs (50 GB, within free tier)$0/mo
Traces (20 GB, within free tier)$0/mo
Total$0/mo ($0/yr)
Winner at 5 engineers: Grafana Cloud at $0/year, vs New Relic at $2,640/year, vs Datadog at $6,768/year. Grafana Cloud is free for small teams — and the gap is enormous.
Head-to-Head: 20-Engineer Team
A 20-engineer team with 30 hosts, 200 GB/month logs, and full APM coverage.
Datadog — 20 Engineers, 30 Hosts
Infrastructure (30 hosts × $15)$450/mo
APM (30 hosts × $40)$1,200/mo
Log Management (200 GB × $0.10 ingestion)$20/mo
Log Indexing (200 GB ≈ 20M events × $1.80)$36/mo
Total$1,706/mo ($20,472/yr)
New Relic — 20 Engineers, 200 GB Logs
Data ingest (200 GB - 100 GB free = 100 GB × $0.25)$25/mo
20 full users × 6 hrs/day × $0.25/hr × 22 days$660/mo
Total$685/mo ($8,220/yr)
Grafana Cloud — 20 Engineers, 200 GB Logs
Metrics (25,000 series, Pro: 15K overage × $8/1K)$120/mo
Logs (150 GB overage × $0.50)$75/mo
Traces (100 GB × $0.50)$50/mo
Pro users (20 × $29)$580/mo
Total$825/mo ($9,900/yr)
Note: At 20 engineers, Grafana Cloud's Pro user seats start to add up. If your team can work with fewer editor seats (3 editors, 17 viewers), Grafana Cloud drops to ~$175/mo — making it the clear winner. The $29/user/mo seat cost is the main lever for Grafana Cloud at scale.
Head-to-Head: 50-Engineer Team
A 50-engineer team with 80 hosts, 500 GB/month logs, and enterprise requirements.
Datadog — 50 Engineers, 80 Hosts
Infrastructure Enterprise (80 × $23)$1,840/mo
APM (80 × $40)$3,200/mo
Log Management (500 GB ingestion)$50/mo
Log Indexing (500 GB ≈ 50M events × $1.80)$90/mo
RUM (500K sessions × $1.50/1K)$750/mo
Total$6,830/mo ($81,960/yr)
New Relic — 50 Engineers, 500 GB Logs
Data ingest (500 GB - 100 GB free = 400 GB × $0.25)$100/mo
50 full users × 6 hrs/day × $0.25/hr × 22 days$1,650/mo
Total$1,750/mo ($21,000/yr)
Grafana Cloud — 50 Engineers, 500 GB Logs
Metrics (80,000 series, Pro: 70K overage × $8/1K)$560/mo
Logs (450 GB overage × $0.50)$225/mo
Traces (300 GB × $0.50)$150/mo
Pro users (50 × $29)$1,450/mo
Total$2,385/mo ($28,620/yr)
Winner at 50 engineers: New Relic at $21K/yr, Grafana Cloud at $28.6K/yr, Datadog at $82K/yr. Datadog is nearly 4x the cost of New Relic and 3x the cost of Grafana Cloud at this scale.
Cost Summary by Team Size
| Team Size | Datadog | New Relic | Grafana Cloud | Cheapest |
|---|---|---|---|---|
| 5 engineers | $6,768/yr | $2,640/yr | $0/yr | Grafana Cloud |
| 20 engineers | $20,472/yr | $8,220/yr | $9,900/yr | New Relic |
| 50 engineers | $81,960/yr | $21,000/yr | $28,620/yr | New Relic |
| 100+ engineers | $150K+/yr | $50K–$80K/yr | $60K–$100K/yr | Negotiate |
Feature Comparison
| Feature | Datadog | New Relic | Grafana Cloud |
|---|---|---|---|
| Infrastructure Monitoring | Best-in-class | Good | Good (Prometheus) |
| APM / Distributed Tracing | Excellent | Good | Good (Tempo) |
| Log Management | Good | Good | Excellent (Loki) |
| Dashboards | Good | Good | Best-in-class |
| Alerting | Excellent | Good | Good |
| AI/ML Features | Watchdog AI | NRAI | Basic |
| Open Source | No | No | Yes (entire stack) |
| Self-Hosted Option | No | No | Yes (free) |
| Vendor Lock-in Risk | High | Medium | Low |
| Ease of Setup | Easiest | Easy | Moderate |
Hidden Costs to Watch
Datadog Hidden Costs
- Custom metrics: $0.05 per custom metric beyond 100/host. Teams with 50 hosts and 200 custom metrics each pay $25/mo extra per host — a 67% surcharge on the base price.
- Container monitoring: Containers count as hosts. A Kubernetes cluster with 20 containers = 20 hosts, not 1.
- Log rehydration: $0.05/GB to search archived logs. Teams that archive to S3 and need to search old data pay twice.
- Annual commit minimums: Enterprise contracts often require $100K+ annual commitments.
New Relic Hidden Costs
- Concurrent user hours: The CCU model makes costs unpredictable. Oncall weeks spike costs 3–5x.
- Data overages: Exceeding your plan's data limit incurs automatic overage charges at $0.30/GB (Standard) or $0.25/GB (Pro).
- Full platform user costs: Every developer who needs APM details needs a full platform user seat — core users can't access tracing data.
Grafana Cloud Hidden Costs
- Pro user seats: At $29/user/mo, editor seats add up fast. A 50-engineer team pays $1,450/mo just in user costs — more than the data ingest costs.
- Metrics series explosion: High-cardinality metrics (e.g., per-request labels) can balloon series counts. One misconfigured label can add thousands of series.
- Self-hosting overhead: If you self-host to save money, factor in 0.5–1 FTE of DevOps time to manage Prometheus, Loki, and Tempo at scale.
Which Platform Should You Choose?
Choose Grafana Cloud if:
- You're a startup or small team (under 15 engineers)
- You want to avoid vendor lock-in (open-source stack)
- You have DevOps expertise to manage the stack
- You want the cheapest possible observability
- Dashboards and log exploration are your primary use case
Choose New Relic if:
- You're a mid-size team (15–50 engineers) with moderate data volume
- You want a per-GB pricing model that's easy to predict
- You need full APM without per-host pricing
- You want 100 GB free data ingest with no strings attached
- You prefer a SaaS platform over managing your own stack
Choose Datadog if:
- You're an enterprise (50+ engineers) with budget for premium tooling
- You need best-in-class APM, infrastructure monitoring, and alerting
- You want AI-powered anomaly detection (Watchdog)
- Ease of setup and minimal configuration are critical
- You're already invested in the Datadog ecosystem
Track Observability Pricing Changes
Datadog, New Relic, and Grafana Cloud all adjust pricing regularly. Get notified when prices change.