Elastic Cloud Hosted is Elastic’s managed service for running Elastic Stack deployments on AWS, Azure and Google Cloud, in which the customer keeps control of hardware configuration, cluster size, node count and versions.[1][3] It is sold in four subscription levels, Standard, Gold, Platinum and Enterprise, which decide the features available, the support level and whether an uptime SLA applies.[1] Billing is resource based: each deployment is charged for the RAM of its nodes over time, plus data transfer, snapshot storage and some usage-based features, paid monthly as you go or prepaid.[2][3]
The same level names apply to an Elastic Cloud organization’s Serverless projects, where they set the support percentage; see Elastic Cloud Serverless pricing.
Editions
| Level | Starting price | Support | Selected additions over the level below | Catalog SKU |
|---|---|---|---|---|
| Standard | USD 99 per month | Limited: web support, 2 contacts, 3-business-day target (platform only) | Elasticsearch, Kibana, Security and Observability basics; AutoOps monitoring | Elastic Cloud Hosted Standard |
| Gold | USD 114 per month | Base: business hours, phone and web, 6 contacts | Reporting, third-party alerting actions, Watcher, multi-stack monitoring | Elastic Cloud Hosted Gold |
| Platinum | USD 131 per month | Enhanced: 24/7/365, 8 contacts, 1-hour urgent target | Advanced security features, machine learning, cross-cluster replication, 99.95% SLA | Elastic Cloud Hosted Platinum |
| Enterprise | USD 184 per month | Premium: 24/7/365, 8 contacts, 30-minute urgent target | Searchable snapshots (cold and frozen tiers), cross-cluster search, Elastic Maps Server, Agent Builder, Workflows, SAML SSO for Elastic Cloud | Elastic Cloud Hosted Enterprise |
All four prices are “as low as” figures “Based on cloud production config, 120 GB storage / 2 zones”, with instance type usage-based pricing.[1] They are reference points for a small deployment, not per-seat or per-node prices.
Self-managed deployments use a different ladder: Free and open (Basic), Platinum for existing customers only, and Enterprise.[5] The self-managed metrics (Resource Units and Billable Nodes) are covered in Elastic licensing.
Metrics
| Elastic name | Catalog row | Notes |
|---|---|---|
| GB of RAM per hour | GB of RAM per hour (Elastic Cloud Hosted) | Deployment capacity; CPU and disk scale with RAM and are included[2] |
Elastic Cloud Hosted billing dimensions are deployment capacity, data transfer, storage, synthetics, inference, workflow executions and Elastic Agent Builder executions.[2] New annual subscriptions bought directly from Elastic are prepaid in Elastic Consumption Units.[4]
Counting / floors
Deployment capacity (as of 2026-09-30). Each node type (Elasticsearch, Kibana, APM, ML and others) is priced per GB of RAM per hour; the cost is the sum over nodes of RAM times running time. Elastic describes this as “purely a function of your current deployment configuration and time”.[2]
Data transfer (as of 2026-09-30). Data in is free. Data out (including traffic through AWS PrivateLink, GCP Private Service Connect or Azure Private Link) and data inter-node are summed across all deployments in the account, with 100 GB free per month; “Once this threshold is passed, a charge is applied for any data transfer used in excess of the 100GB monthly free allowance.” Inter-node charges are waived for Azure deployments.[2] Catalog: Elastic Cloud Hosted: 100 GB monthly free allowance for data out and inter-node transfer.
Snapshot storage (as of 2026-09-30). Snapshot storage is metered hourly and averaged into GB/month over each calendar month across all deployments, at one price in all regions, with 100 GB/month free. Storage API requests are billed per 1,000 calls with 100,000 free per month. Disk for the indices themselves is included in the RAM hours.[2] Catalog: Elastic Cloud Hosted snapshot storage: hourly-averaged GB-month with 100 GB free.
Synthetics (as of 2026-09-30). Browser tests are USD 0.0123 per test run, metered in 60-second increments; lightweight tests are USD 28.00 per location per month per stack for up to 1,000 simultaneous test runs. Tests from private locations incur no execution charge.[1]
Virtualization & partitioning
Not applicable to the customer. Elastic runs Hosted deployments on the chosen cloud provider’s infrastructure; the customer selects hardware profiles and sizes, which drive the RAM-hour charge.[2]
Cloud / BYOL
Hosted is available on AWS, Google Cloud, Azure and Alibaba Cloud, and in FedRAMP regions, across 60 regions.[3] It can also be bought through the AWS, Azure and Google Cloud marketplaces to consolidate the cloud bill.[3] There is no bring-your-own-licence route from a self-managed subscription to Hosted in the pricing documents.
Programs
Uptime SLA (as of 2026-09-30). “99.95% Monthly Uptime SLA for Platinum and Enterprise subscription tiers” applies to Hosted and Serverless; self-managed deployments have no uptime SLA because the customer runs the infrastructure.[3] Catalog: Elastic Cloud: 99.95% monthly uptime SLA only for Platinum and Enterprise.
Changing level. On a monthly subscription the level can be changed from the console: upgrades take effect immediately and downgrades at the next invoice date, and a customer cannot downgrade while using features of its current level.[4] The Monthly, Consumption and Annual commercial programs are described in Elastic licensing.
Out of scope
- Per-region RAM-hour rates in the Elastic Cloud pricing table.
- Elastic Cloud Enterprise and Elastic Cloud on Kubernetes, which are self-managed orchestration products.
- Inference, workflow and Agent Builder rates on Hosted, which reference the Serverless pricing pages.
- The complete feature matrix per level.