Is Elasticsearch free?

The self-managed software has a free basic tier, and you pay for your infrastructure and the time to operate it. Some features, such as RRF merging for hybrid search, are listed under paid subscriptions according to their subscription table, checked on 5 October 2026.

Can I do semantic search in Spanish with Elasticsearch?

Yes. Elastic recommends its ELSER model for English and the multilingual E5 model for other languages. You can also use your own embeddings in a dense_vector field.

Is XEYE suitable for searching logs or long documents?

No. It is designed for lists of short texts, such as products, FAQs or listings. For logs or long documents, Elasticsearch is the right tool.

XEYE vs Elasticsearch: which one to choose

Elasticsearch is a search and analytics engine that you deploy and configure yourself, or buy as a managed service. XEYE is a hosted API that needs no operating. Elasticsearch is the right choice for millions of documents, filters, aggregations or full control. XEYE fits when the catalogue is small and nobody on the team is going to design mappings, embedding pipelines and hybrid ranking.

Updated on October 5, 2026By Joan Martorell

How do they differ?

	XEYE	Elasticsearch
What it is	Hybrid search API over lists of short texts.	Distributed search and analytics engine.
Deployment	Hosted. There is nothing to install.	Self-managed, Elastic Cloud Hosted or Elastic Cloud Serverless.
Semantic search	Included. XEYE manages the model and the embeddings.	`dense_vector` fields with approximate search, and the `semantic_text` type, which generates the embeddings at index time through its inference API.
Models	Two multilingual models to choose from on each training.	Whichever you choose. Elastic recommends its ELSER model for English and E5 for other languages.
Hybrid search	On every query, with nothing to configure.	They recommend merging results with RRF. In self-managed deployments, RRF is listed under the Enterprise subscription; with the free licence you do the merging yourself.
Fuzzy matching	Yes, not configurable.	Yes, with the `fuzzy` query and tuning parameters.
Filters and aggregations	No.	Yes, very complete.
Operations	None.	Self-managed: sizing the cluster and the shards, upgrades, backups and security.
Price	€0.001 per search, no monthly fee.	Self-managed: free at the basic tier, plus your infrastructure. Cloud Hosted: from $99 per month. Serverless: usage-based.
Licence	Closed service.	Source code available under Elastic License 2.0, SSPL 1.0 or AGPLv3.

Elastic data checked in its documentation and on its pricing and subscription pages on 5 October 2026. Check them before deciding.

When is Elasticsearch the better choice?

You have millions of documents, or you also need logs, analytics or aggregations.
You want full control over analyzers, languages, filters, facets and scoring.
The data must stay on your own infrastructure, or you already use Elastic.
You have engineers to operate it, or budget for the managed version.
You want to choose and change the embedding model yourself.

When does XEYE fit?

The catalogue has from hundreds to a few thousand short elements, and a cluster or a managed deployment is out of proportion.
Nobody on the team is going to design mappings, embedding pipelines and hybrid ranking.
You want a cost that follows usage, without an always-on resource.

What do I have to build with Elasticsearch?

Less than a few years ago. The semantic_text type generates the embeddings at index time and splits long texts without you setting up an ingest pipeline. Even so, some decisions remain yours: which model to use, how to combine the text signal with the semantic one, how to size the cluster if you manage it yourself and which subscription tier you need for the features you want.

With XEYE there are none of those decisions. In exchange, you cannot make them either: you do not choose the combination formula or add filters.

What is XEYE missing?

It is not designed for large volumes or for analytics. It has no filters, aggregations, client libraries or service level agreement, and the data is hosted in the service, not on your infrastructure. We have measured it with catalogues of up to 4,727 elements; the data is in measured results.

Sources

Semantic search and hybrid search, in the Elastic documentation.
semantic_text field type and ELSER model.
Elastic subscriptions: which features each tier includes.
Elastic Cloud Hosted pricing and licensing questions.

Frequently asked questions

Is Elasticsearch free?: The self-managed software has a free basic tier, and you pay for your infrastructure and the time to operate it. Some features, such as RRF merging for hybrid search, are listed under paid subscriptions according to their subscription table, checked on 5 October 2026.
Can I do semantic search in Spanish with Elasticsearch?: Yes. Elastic recommends its ELSER model for English and the multilingual E5 model for other languages. You can also use your own embeddings in a dense_vector field.
Is XEYE suitable for searching logs or long documents?: No. It is designed for lists of short texts, such as products, FAQs or listings. For logs or long documents, Elasticsearch is the right tool.

Try it with your own data

Create an account, upload a list and run your first search in about five minutes. You start with €5 of credit, no card required.

Create a free account Read the documentation