Elasticsearch is built for log analytics, not site search.
Elasticsearch is one of the most powerful search and analytics engines ever built. It handles billions of documents across distributed clusters for companies like Netflix and Uber. But if you need search for your website or application, you do not need a JVM, a cluster, or a DevOps team. Sprigr is fully managed, with a client-side mode that runs search in the browser, and zero infrastructure for you to run.
Sprigr vs Elasticsearch for site search
How a purpose-built site search engine compares to an enterprise analytics platform.
| Sprigr | Elasticsearch | |
|---|---|---|
| Built for | Website & app search | Log analytics & full-text search at scale |
| Architecture | Hosted API + client-side mode | Distributed cluster (JVM-based) |
| Search latency | <10ms client-side; tens of ms hosted | 10-500ms (network + cluster processing) |
| Infrastructure required | Yes. None | Servers, JVM, cluster management |
| Ops burden | Zero ops (fully managed) | Shards, replicas, heap tuning, upgrades |
| Pricing | Free tier + flat fee from $49/mo | Self-hosted (free) or Elastic Cloud (~$95+/mo) |
| Per-query costs | Yes. Zero | Compute costs per query |
| Data privacy | Queries local in client-side mode; records hosted per tenant on Cloudflare | Queries processed on server |
| Offline support | Yes. Yes | No. No |
| Setup time | Minutes (REST API + @sprigr/browser) | Hours to days (cluster provisioning) |
| Hybrid semantic search | Yes. Yes (keyword + vector RRF) | Yes. Yes (kNN + BM25) |
| Index size limit | Tens of thousands in client-side mode; up to 500K objects hosted | Billions of records |
| License | Proprietary SaaS | Elastic License 2.0 / SSPL (not OSI-approved) |
What Sprigr does instead
A site search engine built for the actual problem, not the distributed systems problem.
-
Zero infrastructure
No JVM. No cluster. No servers of your own. Sprigr hosts your index on Cloudflare's network, and in client-side mode the browser downloads a compact binary index once and runs the WebAssembly engine locally. You push records to the REST API and add the @sprigr/browser package to your site. There is nothing to provision, patch, or monitor.
-
Sub-10ms latency without a network
Elasticsearch latency depends on cluster load, network distance, and query complexity. In Sprigr's client-side mode, search runs as local computation in the browser and every query returns in under 10 milliseconds, regardless of the user's network conditions, geographic location, or whether they are offline. Hosted queries typically return in tens of milliseconds.
-
Flat pricing with no surprises
Elasticsearch is free to run yourself, but the real cost is infrastructure and engineering time. Elastic Cloud starts around $95 per month for a minimal deployment and scales quickly. Sprigr has a permanent free tier for 1,000 objects and paid plans starting at $49 per month with zero per-query charges.
-
Privacy by architecture
With Elasticsearch, every search query is sent to your server and processed on your infrastructure (or Elastic's cloud). In Sprigr's client-side mode, search queries run locally in the user's browser after a one-time index download, so no query is sent to a server or logged. Your records are hosted on Cloudflare's network, isolated per tenant.
Migrating from Elasticsearch to Sprigr
If you are currently using Elasticsearch for site search, migrating to Sprigr simplifies your stack considerably. The process involves exporting your documents from Elasticsearch and pushing them to the Sprigr REST API.
Elasticsearch stores documents as JSON objects, and Sprigr indexes JSON objects with the same structure. Export your documents from Elasticsearch (using the Scroll API or a simple query), map them to Sprigr's record format (each record needs an objectID and your searchable attributes), and push them via the Sprigr REST API. The API is Algolia-compatible, so if you have experience with Algolia's record format, the structure will be familiar.
On the frontend, you replace your Elasticsearch query logic (whether you were using the JavaScript client, a REST proxy, or a custom backend) with the Sprigr endpoint, or with the @sprigr/browser package in client-side mode, where the index is downloaded once and queries run locally in the browser. You can remove your Elasticsearch cluster, your search proxy server, and any associated monitoring infrastructure.
For teams that built custom relevance tuning in Elasticsearch (boosting, function scores, decay functions), Sprigr provides searchable attribute priority ordering, custom ranking attributes, and hybrid semantic search with reciprocal rank fusion. The tuning model is simpler but covers the common site search scenarios without writing complex query DSL.
What switching saves you
- elasticsearch $95+/mo managed
Elastic Cloud starts at $95/mo. Self-hosted requires DevOps, monitoring, and cluster management.
- sprigr $49/mo flat
No clusters. No DevOps. Hosted search with zero infrastructure of your own. Free tier: 1K objects.
Questions
Is Elasticsearch free?
Elasticsearch was originally Apache 2.0 licensed, but since 2021 it uses the Elastic License 2.0 and SSPL, which are not considered true open-source licenses by the Open Source Initiative. You can download and run it at no software cost, but you are responsible for all infrastructure: servers, storage, memory, networking, monitoring, backups, and ongoing maintenance. Elastic Cloud, the managed hosting option, starts at roughly $95 per month for a minimal production deployment and scales significantly with usage.
Why is Elasticsearch overkill for site search?
Elasticsearch was designed for distributed log analytics, observability pipelines, and searching billions of documents across clusters of machines. Running it for website search with a few thousand pages means provisioning JVM heap memory, configuring shards and replicas, managing cluster health, and handling rolling upgrades. The operational complexity does not match the problem. A site search with 5,000 pages does not need a distributed system.
How does Elastic Cloud pricing compare to Sprigr?
Elastic Cloud pricing is resource-based: you pay for the compute, memory, and storage your cluster consumes. A minimal production deployment with a single availability zone starts around $95 per month and scales into hundreds or thousands per month as you add nodes, RAM, storage, or machine learning features. Sprigr uses flat-fee pricing with no per-query charges. The free tier covers 1,000 objects forever, and paid plans start at $49 per month with zero infrastructure costs to manage.
Can Sprigr handle large datasets like Elasticsearch?
No, and it is not designed to. Elasticsearch handles billions of documents across distributed clusters. Sprigr's hosted plans go up to 500,000 objects, and client-side mode works best for indexes of up to tens of thousands of records. If your dataset has millions of records or requires complex aggregations across terabytes of data, Elasticsearch is the right tool. Sprigr is purpose-built for website and application search where the dataset is manageable and the priorities are simplicity, speed, and zero operational overhead.
What about OpenSearch as an alternative?
OpenSearch is the AWS-maintained fork of Elasticsearch, created after Elastic changed its license in 2021. It is Apache 2.0 licensed and genuinely open source. However, it shares the same fundamental architecture and operational complexity: JVM tuning, cluster management, shard allocation, rolling upgrades, and monitoring. If your concern with Elasticsearch is licensing, OpenSearch is a solid alternative. If your concern is operational overhead for a simple site search use case, neither Elasticsearch nor OpenSearch reduces that burden. Sprigr eliminates that infrastructure entirely: it is fully managed, and its client-side mode can even run search in the browser.
Site search without the infrastructure.
No servers, no clusters, no JVM.