Grafana Cloud is Grafana Labs' hosted observability stack: Grafana dashboards over managed metrics, logs, traces and profiles, plus add-ons for frontend monitoring, synthetics, load testing and incident response. It has a free tier, a pay-as-you-go Pro plan and an Enterprise contract.
This guide covers the plans, how each meter bills, the charges that surprise teams, three worked monthly bills and how to cut one.
Maple publishes this guide and competes with Grafana Cloud. Section 06 compares the two; every Grafana Cloud figure comes from Grafana Cloud's own pages, linked at the end.
Key takeaways
- Free includes 10,000 metric series, 50 GB each of logs and traces, 3 users and 14 days of retention.
- Pro is $19 a month plus usage above the same allowance, with 30 days for logs and traces and 13 months for metrics.
- Metrics are $6.50 per 1,000 active series. Scraping faster than once a minute multiplies that.
- Logs and traces are $0.05 per GB to process and $0.40 per GB to write. Keeping them past 30 days adds $0.10 per GB per 30 days.
- Every user who logs in during a month counts. Past 3, each is $8.
01 · Plans
Grafana Cloud plans and prices
| Per month | Free | Pro | Enterprise |
|---|---|---|---|
| Price | $0 | $19 + usage | From $25,000 a year |
| Metrics | 10k series | 10k, then $6.50 per 1k | From $3 per 1k |
| Logs | 50 GB | $0.05 per GB processed, 50 GB written, then $0.40 per GB | Custom |
| Traces | 50 GB | $0.05 per GB processed, 50 GB written, then $0.40 per GB | Custom |
| Profiles | 50 GB | $0.05 per GB processed, 50 GB written, then $0.40 per GB | Custom |
| Active users | 3 | 3, then $8 each | Custom |
| Frontend Observability | 50k sessions | 50k, then $0.75 per 1k | Custom |
| Application Observability | 2,232 host-hours | $0.025 per host-hour, none included | Custom |
| Kubernetes Monitoring | 2,232 host-hours, 37,944 container-hours | Then $0.01 per host-hour, $0.0007 per container-hour | Custom |
| Synthetics | 100k API, 10k browser runs | Then $5 per 10k API, $50 per 10k browser | Custom |
| IRM | 3 active users | Then $20 each | Custom |
| Retention | 14 days | 13 months metrics, 30 days logs and traces | Custom |
Pro includes the free allowance every month and bills what goes past it. Application Observability is $18 per host at a full month of hours; Kubernetes Monitoring is $7.20 per host and $0.50 per container.
The free tier
The free tier costs nothing with no credit card and no end date. It runs the same services as Pro within usage limits, with 14 days of retention and community support. Pro adds the longer retention and email support, and turns the limits into an allowance with usage billed past it. Enterprise starts at a $25,000 a year spend commitment with custom retention.
Verified 2026-11-21 against grafana.com/pricing and Grafana's invoice docs for metrics and logs.
02 · Meters
How each Grafana Cloud meter bills
Metrics: active series and DPM
An active series is a time series that received a data point in the last 20 minutes. Grafana measures active series through the month and bills the 95th percentile, which forgives the busiest 5% of the month, about 36 hours. A stack that normally runs 6,000 series and spikes to 30,000 for a day is still billed for 6,000.
DPM is data points per minute per series. Pro includes 1 DPM, a 60-second scrape. The invoice docs give the rule as usage = max(active_series, total_dpm / included_dpm). At a 30-second scrape, 50,000 series send 100,000 data points a minute and bill as 100,000 series. A 15-second scrape, common in Prometheus configs, is 4 DPM.
Logs and traces: process, write, retain
Logs, traces and profiles bill per GB in three parts on Pro: $0.05 to process, $0.40 to write and $0.10 to retain. The 50 GB allowance offsets the write charge only, so processing is billed from the first GB. Process is measured on everything Grafana receives. Write is measured on what is kept after Adaptive Telemetry features such as Adaptive Logs drop data, so dropping cuts the write charge but not the process charge.
When retain applies
The pricing page lists retain as a third per-GB rate on Pro. The logs invoice docs bill it "for each additional retention increment beyond the minimum": paid accounts keep 30 days at no retain charge, and each further 30 days costs $0.10 per GB. The worked bills below use the default 30 days, so they carry no retain charge.
Queries are included up to a fair-use ratio: 100 times your monthly written log volume, then $0.003 per GB scanned. Metrics queries are included up to 20 billion query units per 1,000 billable series.
Users
Grafana defines an active user as anyone who logs into Grafana during the billing month. Three are included; each one after that is $8 a month, or $55 with Enterprise plugins, a paid add-on. Incident response (IRM) counts its own users at $20 each past 3.
Volume discounts
Pro rates step down automatically as volume grows. These are the published bands:
| Meter | First band | Second band | Third band |
|---|---|---|---|
| Series, per 1k | $6.50 to 100k | $5.90 to 200k | $5.50 above |
| Process, per GB | $0.05 to 10,000 GB | $0.046 to 25,000 GB | $0.044 above |
| Write, per GB | $0.40 to 1,000 GB | $0.37 to 2,500 GB | $0.35 above |
| Retain, per GB | $0.10 to 5,000 GB | $0.092 to 12,500 GB | $0.088 above |
| Active users | $8 to 25 | $7.25 to 50 | $6.80 above |
The worked examples below stay inside the first band, so the band rules do not change their totals.
03 · Surprises
What surprises teams on Grafana Cloud
- The scrape interval is a price multiplier. Moving from 60 to 15 seconds bills four times the series, with no new metric added.
- Labels create series. Every unique label combination is a series. A
user_idorpodlabel on a busy metric can add tens of thousands. - Anyone who logs in is a user. A product manager who opens one dashboard in a month costs $8 like an engineer on call.
- Curated apps add host-hours. Application Observability bills $0.025 per host-hour on top of the traces it reads, and Kubernetes Monitoring bills hosts and containers. Grafana's FAQ says you can opt out of the curated experience and pay only for the telemetry.
- Dropped logs still pay to process. Adaptive Logs lowers the $0.40 write charge, not the $0.05 process charge.
- Free retention is 14 days. Teams that start free and need a month of logs move to Pro.
04 · Examples
What a month of Grafana Cloud costs
Three bills on Pro at the prices checked 2026-11-21, in a 30-day month. Each bar splits one bill by category:
Small team: $21.50 a month
Three engineers, 8,000 active series at a 60-second scrape, 30 GB of logs and 20 GB of traces. Every write fits the free allowance, so on Pro they pay the $19.00 platform fee plus processing on all 50 GB, $21.50 in total, for 30 days of logs and traces instead of 14. On the free tier the same team pays nothing.
Mid-sized team: $741.00 a month
Twelve people log in during the month. They run 40,000 active series at a 60-second scrape, send 400 GB of logs and 300 GB of traces, and use Application Observability on 10 hosts all month.
| Line | Calculation | Per month |
|---|---|---|
| Pro platform fee | Monthly | $19.00 |
| Active users | 12 − 3 included = 9 × $8 | $72.00 |
| Metrics | 40k − 10k included = 30k × $6.50 | $195.00 |
| Logs | 400 GB × $0.05 process + max(0, 400 − 50 included) = 350 GB × $0.40 write | $160.00 |
| Traces | 300 GB × $0.05 process + max(0, 300 − 50 included) = 250 GB × $0.40 write | $115.00 |
| Application Observability | 10 hosts × 720 h = 7,200 host-hours × $0.025 | $180.00 |
| Total | $741.00 |
Logs and traces are the largest share. Keeping them 60 days instead of 30 adds one retain increment, about $70.00, for a total of $811.00.
Same team, 30-second scrape: $1,001.00 a month
The team halves the scrape interval to see shorter spikes. The 40,000 series now send 2 data points a minute each, so they bill as 80,000:
| Line | Calculation | Per month |
|---|---|---|
| Metrics, 60 s | 40k − 10k included = 30k × $6.50 | $195.00 |
| Metrics, 30 s | 40k series × 2 DPM = 80k − 10k included = 70k × $6.50 | $455.00 |
| Bill change | +$260.00 |
No series, label or dashboard changed. A 15-second scrape would bill the same 40,000 series as 160,000, which crosses into the second volume band.
05 · Cutting the bill
How to cut a Grafana Cloud bill
- 01
Read the cost dashboards
Grafana Cloud's Cost Management dashboards show month-to-date usage, refreshed hourly. Start with the meter that grew most since last month.
- 02
Scrape once a minute
Pro includes 1 DPM. Keep a 60-second interval unless a dashboard needs finer resolution, and set it per job rather than globally.
- 03
Turn on Adaptive Metrics
It identifies unused series and recommends aggregating them. Grafana says it can cut metrics by up to 80%; your number depends on how many unused labels you send.
- 04
Drop log patterns nobody reads
Adaptive Logs recommends patterns to drop, which saves the $0.40 write charge on each GB. Dropping in your Collector or agent before sending saves the $0.05 process charge too.
- 05
Audit who logs in
Each person who logs in during a month counts. Remove accounts nobody uses and keep Enterprise plugins to the people who need them, since those users cost $55 instead of $8.
- 06
Check the host-hour products
Application Observability and Kubernetes Monitoring bill per host-hour for the curated views. If the team only uses the raw metrics and traces, opt out and pay for the telemetry alone.
06 · Maple
Maple pricing for the same workload
Maple is an OpenTelemetry observability platform. It has one plan at $39/month with 100 GB each of logs, traces and metrics included, then $0.30 per GB for each signal. There is no per-user, per-series or per-host charge. Logs and traces are kept 30 days and metrics 90.
The mid-sized team on Maple, which meters uncompressed OTLP bytes, with metrics estimated at 0.1 KB per data point. At 1 DPM, 40,000 series send 1,728,000,000 data points a month:
| Line | Calculation | Per month |
|---|---|---|
| Plan | Monthly price | $39.00 |
| Logs | 400 GB − 100 included = 300 GB × $0.30 | $90.00 |
| Traces | 300 GB − 100 included = 200 GB × $0.30 | $60.00 |
| Metrics | 172.8 GB − 100 included = 72.8 GB × $0.30 | $21.84 |
| Total | $210.84 |
If metric points average 0.3 KB instead of 0.1, the total is $314.52. At a 30-second scrape Maple meters twice the metric bytes and the 0.1 KB total is $262.68. Scrape interval moves both bills; on Maple it moves the bytes, on Grafana Cloud the billed series.
Grafana Cloud is the better fit in several cases. Its free tier covers a small team at $0, and Maple has no free cloud plan. Pro keeps metrics for 13 months against Maple's 90 days. Grafana dashboards can read from many data sources outside Grafana Cloud, and the Grafana stack can be self-hosted from open-source components. Maple fits teams that send OpenTelemetry and want one store for all three signals, no per-user charge, and source code they can run themselves.
Maple
Price your own month
The calculator prices Grafana Cloud and Maple on the same sliders, and the compare page goes feature by feature.
Sources and further reading