Compare / Elastic Cloud
WarpSearch vs Elastic Cloud Hosted (managed Elasticsearch)
Elastic's managed Elasticsearch deployments on AWS, Azure and Google Cloud. You pick a hardware profile, node sizes and zones, and pay per GB of RAM per hour for every node, whether or not anyone is querying.
Elastic Cloud pricing for hosted Elasticsearch is per GB of RAM per hour, for every node and zone you run. WarpSearch answers the Elasticsearch REST API with no deployment to size. Below is what managed Elasticsearch on Elastic Cloud Hosted costs for four real workloads, next to WarpSearch; for Elastic's serverless option, see Elasticsearch Serverless pricing.
The scoreboard
Price
Cheaper on 3 workloads
Monthly, list prices. Bars are drawn to the middle of a range.
How they bill
| WarpSearch | Elastic Cloud | |
|---|---|---|
| Billing unit | Per GB stored, per GB written, per request | Per GB of RAM per hour for each node and zone, plus snapshot storage ($0.033/GB-month) and data transfer ($0.05/GB out, $0.016/GB between nodes) |
| Smallest HA deployment | $10 a month minimum | "As low as" $99 a month (Standard, 120 GB storage across 2 zones) |
| Subscription tiers | One rate card | Standard, Gold, Platinum and Enterprise. The same node costs up to 2.3x more on Enterprise |
| Scale to zero | Yes, after 15 minutes idle | No. Nodes run all month |
| Capacity planning | None | You pick the hardware profile, RAM per node, zones and shards |
| Log retention | One index, storage tiered per GB | Hot nodes plus frozen nodes over searchable snapshots. Frozen needs Enterprise; each GB of frozen RAM addresses about 1.6 TB |
What Elastic Cloud cannot do, or does only in part: Provisioned deployments billed per GB of RAM per hour, per zone, around the clock. CPU and disk scale with the RAM you buy. Frozen and cold tiers (searchable snapshots) need the Enterprise subscription. Deployments with 6 or more Elasticsearch nodes add dedicated master nodes.
Price
Monthly cost at list prices in US East.
1M text chunks with 768-dim embeddings, 7 GB of JSON. 1M queries a month spread through the day. 1 GB written a month.
| WarpSearch | Elastic Cloud | |
|---|---|---|
| Storage | $0.56 (7 GB held x $0.08) | $0 (snapshots fit the free 100 GB) |
| Writes | $1.00 (1 GB x $1.00) | In node cost |
| Search | $2.00 (1M requests) | $64 to $256 (1 x 4 GB vector search node with no HA, or 2 x 8 GB across two zones, at $0.0219/GB-hour) |
| Data returned | $0 (about 40 GB, inside the free 100 GB) | $0 (40 GB, inside the free 100 GB) |
| Minimum | $10 minimum applies (usage is $3.56) | Kibana, about $29 (1 GB at $0.0396/GB-hour) |
| Total a month | $10 | $93 to $285 |
Our sizing, Standard subscription. Quantized 768-dim vectors for 1M chunks fit a 4 GB node; full float vectors in memory need 8 GB per copy.
5M products with 384-dim embeddings, 10 GB of JSON. 50M queries a month (about 19 a second on average). 20 GB written a month.
| WarpSearch | Elastic Cloud | |
|---|---|---|
| Storage | $0.60 (7.5 GB held x $0.08) | $0 (snapshots fit the free 100 GB) |
| Writes | $20 (20 GB x $1.00) | In node cost |
| Search | $100 (50M requests) | $480 to $1,463 (2 x 15 GB vector search nodes at $0.0219, or 2 x 15 GB CPU optimized nodes at $0.0668) |
| Data returned | $18 (250 GB returned, 100 GB free, before compression) | $8 (250 GB out and 20 GB replication, less the free 100 GB) |
| Minimum | Cleared | Kibana, about $29 |
| Total a month | $139 | $516 to $1,500 |
Our sizing, Standard subscription. A 15 GB vector search node has about 2 vCPUs, a CPU optimized one about 8; the right size depends on your peak query rate, which varies from the 19 a second average.
Performance
| Test | Result |
|---|---|
| Keyword search, 2.8M docs (LoTTE) | 2.1 ms median, 3,700 queries/s |
| Keyword search, 10M passages (MS MARCO v2) | 2.4 ms median (5.4 ms with an exact total hit count) |
| Keyword search, small BEIR sets | 1 to 6 ms median |
| Vector search, 10M x 128-dim (BigANN SIFT) | 0.964 recall@10 at 6.2 ms median, 162 queries/s |
| Vector search, 1M x 1536-dim (arXiv) | 0.961 recall@10 at 15.7 ms median |
| Filtered vector search, 10M (YFCC) | 0.959 recall@10 at 10.4 ms median |
| Log filters, 22M events (http_logs) | 1 to 2 ms median |
| Log aggregations, 22M events | 190 to 430 ms median (sort, daily histogram, avg by status) |
| Ingest, short text | 63,000 to 97,000 docs/s |
| Ingest, logs (247M events) | 76,000 to 81,000 events/s |
| Ingest, 768 to 1536-dim vectors | 2,200 to 4,000 vectors/s |
| Time to searchable | Seconds, or immediate with commit=true |
| Cold start | Not measured yet. First query on a 2.3 GB index: 17 to 19 ms vs 2.7 ms warm (page cache only) |
Measured October 2026 through our Elasticsearch-compatible API, warm cache, client in the same region, single runs. Every test ran on deliberately minimal hardware, 2 to 4 CPUs (a 4 CPU container or an AWS i8g.large with 2 vCPUs), so these are the least favorable results we can produce. More hardware only improves them. Numbers exclude internet round trips.
Bottom line
Choose WarpSearch for
- Search apps: $10 vs $93 to $285 for RAG, and $139 vs $516 to $1,500 for the catalog workload.
- Large logs: $39,060 vs about $69,400 or more, with no Enterprise subscription needed for long retention.
- No hardware profiles, RAM sizes, zones or shards to size, and no paying for idle hours.
WarpSearch costs a fraction of a hosted Elasticsearch deployment for apps and catalogs, with no nodes, zones or subscription tiers to pick.
Questions
How much does Elastic Cloud cost?
Elastic Cloud Hosted lists "as low as" $99 a month on Standard, $114 on Gold, $131 on Platinum and $184 on Enterprise, for 120 GB of storage across 2 zones. Real deployments bill per GB of RAM per hour: $0.0219 to $0.0668 on Standard in AWS us-east-1, depending on the hardware profile.
Is Elasticsearch free or paid?
Self-managed Elasticsearch is free to download and run, but you pay for the servers and run them yourself. Hosted Elasticsearch on Elastic Cloud is paid, billed by RAM-hours. Features such as frozen-tier searchable snapshots need the Enterprise subscription, which costs about 2.3x Standard per GB-hour on vector search hardware.
What is managed Elasticsearch?
A provider runs Elasticsearch deployments for you. On Elastic Cloud Hosted you still choose the hardware profile, node sizes and number of zones, and you pay for every node-hour. Elastic recommends at least two zones for production, which doubles the data nodes you pay for.
Elastic Cloud Hosted or Elasticsearch Serverless?
Hosted bills RAM-hours for nodes you size. Elasticsearch Serverless removes node sizing but bills VCU-hours, about $100 to $210 a month for a small search app. WarpSearch bills per GB stored, per GB written and per request, with a $10 monthly minimum.
What does hosted Elasticsearch cost at petabyte scale?
Keeping 1 PB of logs searchable on Elastic Cloud Hosted, with a week on hot nodes and the rest in the frozen tier on the Enterprise subscription, is about $69,400 a month or more. WarpSearch is $39,060 for the same workload, with no subscription tier needed for long retention.
Sources
- Elastic Cloud Hosted pricing
- Elastic Cloud pricing table (AWS us-east-1)
- Hosted deployment billing dimensions
- AWS instance configurations
- Plan for production (zones, dedicated masters)
- Trail of Bits benchmark
- WarpSearch pricing
Rates are on-demand list prices from Elastic's pricing table for AWS us-east-1 (valid from May 1, 2026), read on October 22, 2026, at 730 hours a month. Deployment sizes are our estimates, not Elastic sizing guidance. Annual commitments and marketplace deals lower the price. Prices checked October 20, 2026. List prices in US dollars, before discounts. Estimates are marked. Tell us if something is out of date: contact us.