Grafana Cloud Pricing 2026: Free Tier Limits, Pro Costs, and Where You'll Hit the Bill
Grafana Cloud's free tier is the most generous in observability. But costs compound fast once you cross the limits on metrics, logs, and traces. Here's exactly what you pay in 2026.
Grafana Cloud is the managed SaaS version of the Grafana observability stack: Prometheus for metrics, Loki for logs, Tempo for traces, and Grafana for dashboards and alerting. The free tier is the most generous in the observability market. The metered pricing beyond it is straightforward - until you understand how quickly metric series, log volume, and trace storage multiply in real applications.
This guide covers every billing dimension in Grafana Cloud in 2026 and where costs compound unexpectedly.
How Grafana Cloud Pricing Works
Grafana Cloud uses a consumption model with three components:
- Active metric series - billed per 1,000 series per month
- Log volume - billed per GB ingested
- Trace volume - billed per GB stored
Each dimension has a free tier allocation. Beyond those allocations, you pay metered rates. There is no single flat monthly price.
Free Tier - What's Included
| Resource | Free allocation |
|---|---|
| Active metric series | 10,000 series |
| Logs | 50 GB/month |
| Traces | 50 GB/month |
| Profiles | 50 GB/month |
| k6 load testing | 500 VUh/month |
| Alerting | Unlimited rules |
| Dashboards | Unlimited |
| Users | 3 |
| Data retention (metrics) | 13 months |
| Data retention (logs/traces) | 14 days |
The free tier allocations are reset monthly and do not roll over. You can add a credit card without triggering charges - Grafana only bills when you cross the free limits.
What the free tier doesn't include: Grafana Cloud Synthetic Monitoring (uptime checks) is not in the free tier. It's a separate product with per-execution pricing. A dedicated uptime monitoring tool is cheaper for teams that primarily need availability alerts.
Pro Plan - Metered Pricing Rates
| Resource | Free | Pro metered rate |
|---|---|---|
| Active metric series | 10,000 included | $8 / 1,000 series / month |
| Logs | 50 GB included | $0.50 / GB |
| Traces | 50 GB included | $0.50 / GB |
| Profiles | 50 GB included | $0.50 / GB |
| k6 VUh | 500 included | $0.001 / VUh |
| Additional users | 3 included | $8 / user / month |
| Synthetic monitoring | Not included | $0.01 / 1,000 executions |
Total Cost Examples
Small team: microservices application with 50,000 active metric series, 200 GB logs/month
| Line item | Calculation | Monthly cost |
|---|---|---|
| Metric series | (50,000 − 10,000) ÷ 1,000 × $8 | $320 |
| Logs | (200 − 50) GB × $0.50 | $75 |
| Traces | Under 50 GB | $0 |
| Users | 5 users, 2 above free | $16 |
| Total | $411/month |
Mid-size team: 200,000 active metric series, 1 TB logs/month, 500 GB traces/month
| Line item | Calculation | Monthly cost |
|---|---|---|
| Metric series | (200,000 − 10,000) ÷ 1,000 × $8 | $1,520 |
| Logs | (1,000 − 50) GB × $0.50 | $475 |
| Traces | (500 − 50) GB × $0.50 | $225 |
| Users | 10 users, 7 above free | $56 |
| Total | $2,276/month |
The metric series line item is the biggest driver in most bills. Cardinality explosion - metric series multiplying because labels include high-cardinality values like user IDs or request IDs - is the most common cause of unexpected Grafana Cloud costs.
What Causes Cardinality Explosion
Active metric series count is determined by unique label combinations. A metric http_requests_total with labels {method, status, endpoint} across a service with 20 endpoints generates roughly 60 series (3 methods × 4 status codes × 5 endpoints). Add a user_id label and that same metric generates millions of series.
Common cardinality mistakes:
- Labels with unbounded values - user IDs, session IDs, request UUIDs
- Over-labeling with infrastructure metadata - pod names in Kubernetes that rotate frequently
- Importing third-party exporters without filtering - many default configurations emit far more metrics than you need
Grafana Cloud's UI shows active series count per metric. Audit high-cardinality metrics before your bill scales past expectations.
Grafana Cloud vs. Self-Hosting
Self-hosted Grafana (OSS) eliminates the Grafana Cloud bill entirely. You pay for the infrastructure running Prometheus, Loki, Tempo, and Grafana instead.
| Factor | Grafana Cloud | Self-hosted |
|---|---|---|
| Infrastructure cost | $0 (covered by metered rates) | $50–$500/month depending on scale |
| Maintenance burden | None | High - you manage upgrades, storage, HA |
| Scaling | Automatic | Manual |
| Long-term retention | Included | Additional storage cost |
| Setup time | Minutes | Hours to days |
For small teams, Grafana Cloud's free tier beats self-hosting on total cost. At large scale (200,000+ metric series, multi-TB log volumes), self-hosting becomes cheaper - but requires engineering time to operate.
Grafana Cloud Synthetic Monitoring Pricing
Grafana's synthetic monitoring product runs HTTP, TCP, DNS, and ICMP checks from 20+ probe locations. It is priced separately from the core metrics/logs/traces model:
- Free: 100,000 check executions per month (roughly 2 checks per minute across all monitors combined)
- Metered: $0.01 per 1,000 executions beyond the free tier
A single HTTP monitor checking every minute uses 43,200 executions per month. The free 100,000 execution budget covers about 2 monitors at 1-minute intervals. For teams with more than a handful of endpoints to monitor, the execution costs add up.
Dedicated uptime monitoring tools handle this use case at lower cost with more built-in features. See how to choose an uptime monitoring tool for a full comparison of approaches.
Grafana Cloud vs. Alternatives
| Tool | Use case | Free tier | Starting price | Multi-region consensus |
|---|---|---|---|---|
| Grafana Cloud | Full observability stack | 10k metric series, 50 GB logs | Metered, ~$8/1k series | Probe locations, no consensus |
| Datadog | Full observability stack | 5 hosts, 1-day retention | $15/host/month | Yes |
| New Relic | Full observability stack | 100 GB data/month | $0 (user-based) | Yes |
| Vantaj | Uptime and availability monitoring | 20 monitors | $9/month | Yes (default) |
| UptimeRobot | Basic uptime monitoring | 50 monitors | $7/month | No |
Grafana Cloud competes with Datadog and New Relic for observability platform coverage - not with dedicated uptime monitoring tools. If your primary need is knowing when a service goes down, a dedicated uptime tool is simpler and cheaper.
Grafana Cloud Advanced and Enterprise Pricing
Grafana Cloud Advanced and Enterprise plans add:
- Enhanced SLAs (99.5%+ uptime SLA vs. best-effort on free/Pro)
- SAML/SSO for user authentication
- Private data source connections (no public internet exposure)
- Dedicated account management
- Volume discounts on metered usage
Pricing for Advanced and Enterprise is negotiated with a sales team. There is no public rate card. Teams at $2,000+/month on Pro typically qualify for annual contract pricing with 15 to 20% discounts.
The Bottom Line
Grafana Cloud's free tier is the best entry point in observability - 10,000 metric series, 50 GB of logs, and unlimited alerting at no cost. The metered pricing is transparent and predictable once you understand your cardinality and log volume.
For teams that primarily need uptime monitoring, Grafana Cloud Synthetic Monitoring is more expensive per check than dedicated tools. Vantaj covers HTTP, SSL, DNS, heartbeats, and status pages at a flat rate that doesn't scale with check frequency.
Related Guides
How we tested and compared tools
We use one scoring model across comparison articles to keep recommendations consistent.
Test window: Last 30 days before publish date
Uptime check interval: 60-second checks
Alert channels tested: Email, Slack, Webhook
Pricing last checked: July 3, 2026
Criteria and weights
- Reliability and alert quality: 40%
- Setup and daily usability: 25%
- Integrations and coverage: 20%
- Pricing clarity and value: 15%
Sample checks
- Homepage HTTP check from multiple regions
- SSL certificate expiry monitoring
- DNS resolution and nameserver checks
- On-call and escalation flow validation
Known limitations
- Enterprise contract pricing is often private
- Vendors change limits and bundles without notice
- Some findings depend on the selected region and plan tier
Data sources
- Official vendor docs and changelogs
- Public pricing pages
- Hands-on setup and test runs by Vantaj team
Ready to try Vantaj?
Start monitoring in under 60 seconds. No credit card required.