This quick start shows you how to index documents stored as Redis hashes and run search queries against the index. To learn how to index and search JSON documents, see the Search JSON quick start.


For this tutorial, you need:

RediSearch with redis-cli

To begin, connect to your database with redis-cli.

Create an index

The FT.CREATE command lets you create an index. It indexes Redis hashes by default. However, as of RediSearch v2.2, you can also index JSON documents if you have the RedisJSON module enabled.

To set the language used for indexing and stemming, include the optional LANGUAGE keyword. If you do not specify a language, it defaults to English. See Supported languages for additional language options.

When you define an index, you also need to define the schema, or structure, of the data you want to add to the index.

The schema for this example has four fields:

  • title (TEXT)
  • body (TEXT)
  • url (TEXT)
  • visits (NUMERIC)

To define the schema and create the index, run:> FT.CREATE database_idx PREFIX 1 "doc:" LANGUAGE english SCORE 0.5 SCORE_FIELD "doc_score" SCHEMA title TEXT body TEXT url TEXT visits NUMERIC

This command indexes all hashes with the prefix “doc:”. It will also index any future hashes that have this prefix.

By default, all documents have a score of 1.0. However, you can configure an index’s default document score with the SCORE <default_score> option.

If you want to allow document-specific scores, use the SCORE_FIELD <score_attribute> option.

Add documents

After you create an index, you can add documents to the index.

Use the HSET command to add a hash with the key “doc:1” and the fields:> HSET doc:1 title "RediSearch" body "RediSearch is a powerful indexing, querying, and full-text search engine for Redis" url "<>" visits 108
(integer) 4

If you want to make a document appear higher or lower in the search results, add a document-specific score.

To do so, use the score_attribute set by the SCORE_FIELD option during index creation (score_attribute is doc_score in this case):> HSET doc:2 title "Redis" body "Modules" url "<>" visits 102 doc_score 0.8
(integer) 5

Search an index

Use FT.SEARCH to search the index for any documents that contain the words “search engine”:> FT.SEARCH database_idx "search engine" LIMIT 0 10
1) (integer) 1
2) "doc:1"
3) 1) "title"
   2) "RediSearch"
   3) "body"
   4) "RediSearch is a powerful indexing, querying, and full-text search engine for Redis"
   5) "url"
   6) "<>"
   7) "visits"
   8) "108"

Drop an index

To remove the index without deleting any associated documents, run the FT.DROPINDEX command:> FT.DROPINDEX database_idx


You can use RediSearch suggestion commands to implement auto-complete.

Active-Active databases do not support RediSearch suggestions.

Use FT.SUGADD to add a phrase for the search engine to suggest during auto-completion:> FT.SUGADD autocomplete "primary and caching" 100
"(integer)" 1

Test auto-complete suggestions with FT.SUGGET:> FT.SUGGET autocomplete "pri"
1) "primary and caching"

RediSearch with Python

If you want to use RediSearch within an application, these client libraries are available.

The following example uses the Redis Python client library redis-py, which supports RediSearch commands as of v4.0.0.

This Python code creates an index, adds documents to the index, displays search results, and then deletes the index:

import redis
from import TextField, NumericField
from import IndexDefinition
from import Query

# Connect to a database
r = redis.Redis(host="<endpoint>", port="<port>", 

# Options for index creation
index_def = IndexDefinition(
                prefix = ["py_doc:"],
                score = 0.5,
                score_field = "doc_score"

# Schema definition
schema = ( TextField("title"),

# Create an index and pass in the schema
r.ft("py_idx").create_index(schema, definition = index_def)

# A dictionary that represents a document
doc1 = { "title": "RediSearch",
         "body": "RediSearch is a powerful indexing, querying, and full-text search engine for Redis",
         "url": "<>",
         "visits": 108

doc2 = { "title": "Redis",
         "body": "Modules",
         "url": "<>",
         "visits": 102,
         "doc_score": 0.8

# Add documents to the database and index them
r.hset("py_doc:1", mapping = doc1)
r.hset("py_doc:2", mapping = doc2)

# Search the index for a string; paging limits the search results to 10
result = r.ft("py_idx").search(Query("search engine").paging(0, 10))

# The result has the total number of search results and a list of documents

# Delete the index; set delete_documents to True to delete indexed documents as well

Example output:

$ python3 
[Document {'id': 'py_doc:1', 'payload': None, 'visits': '108', 'title': 'RediSearch', 'body': 'RediSearch is a powerful indexing, querying, and full-text search engine for Redis', 'url': '<>'}]

More info