Packages and Index Setup by API - LangChain, LlamaIndex, MCP

An index ready for site search, then the framework you use

Packages and index setup

Get an index ready for site search without opening the Control Panel, then plug it into the framework you already use. The packages do the setup checks for you and call the same endpoints as the rest of this section.

Set up the index

These calls go to https://opensolr.com.

vector_regions

The regions where site search, search by meaning and the chat bot run, with the one nearest to you marked when you ask.

Reference
get_env

The regions and servers you can create an index on; with web_crawler=true, only the ones with site search.

Reference
create_index

Create an Opensolr Index in the region you choose.

Reference
check_schema

Tell whether an index is ready for pushed data and search by meaning, and exactly why not.

Check and install the schema
apply_schema

Install the Opensolr site search schema on an index.

Check and install the schema
get_account_summary

Signed: your plan, its limits and what you used, AI requests included, in one call.

Reference

Packages

LangChain (Python)

A vector store with the vectors made by Opensolr and hybrid search, nothing to run on your side.

LangChain
LlamaIndex (Python)

A vector store and an embedding model for LlamaIndex, on your Opensolr Index.

LlamaIndex
Haystack (Python)

A document store and a hybrid retriever for Haystack pipelines.

Haystack
MCP server

Search, indexing of files, reading of pictures and answers as tools any AI agent can call.

MCP
Laravel Scout (PHP)

A Scout engine: your Laravel models searched by words and by meaning.

Laravel Scout

Each package checks an index once before using it and says exactly what it cannot do; each one also offers the schema check and install. More in Integrations.

Site search API pages

Back to Enterprise Site Search