Skip to content

redis-search-django

Index Django models into Redis Query Engine and query them with the lookups you already know — keyword search, facets, and nearest neighbors.

Get started Run the demo

  • First index in minutes

    Declare a Document, run redisearch create + populate, and search with objects.filter().

    Getting started

  • Django-style query

    Q, lookups, order_by, stock Paginator, and to_queryset() that keeps RediSearch ranking.

    Query API

  • Click-through demo

    A catalog app with search, KNN, aggregations, async, and CRUD.

    Demo app

  • Stays in sync

    Signals index create / update / delete — including related FK, O2O, and M2M changes.

    Signals

How it fits together

flowchart LR
    A[Django models] -->|signals / populate| B[Document]
    B --> C[Redis JSON + FT.SEARCH]
    D[Views / services] -->|Document.objects| C
    D -->|facets / Aggregate / knn| C

Django stays the source of truth. Redis holds a search copy of the fields you declare. Live save() / delete() keep that copy current; the redisearch command covers first setup, schema changes, and bulk imports.

What you get

Need API
Keyword + filters Document.objects.filter(name__search="shoes", price__lte=150)
Facets .facets("category__name", "tags__name")
Stats .aggregate(Aggregate().group_by("vendor__name").avg("price", "avg"))
Meaning .knn("comfortable running shoes", k=10)
Django rows .to_queryset() — same order as Redis
ASGI acount, aget, ato_queryset, async for

Requirements

Runtime Versions
Python 3.10 – 3.15 (free-threaded from 3.15t)
Django 5.2, 6.0, 6.1
redis-py ≥ 8.0 (installed automatically)
Redis 8+ with Query Engine and RedisJSON

Do not import redis in application code. Types you need are re-exported from redis_search_django.

Next