Meilisearch Python

Meilisearch Python is the official API client that lets Python applications talk to a Meilisearch instance. It wraps the full Meilisearch REST API — indexes, documents, search, settings, and tasks — behind an ergonomic, typed client so you can add fast, typo-tolerant, relevance-ranked search to any Python project.

The library handles authentication, request serialization, and response parsing for you, exposing simple Client and Index objects. It supports keyword search, filtering, faceting, and hybrid semantic search, and stays compatible with Meilisearch v1.2 and above.

What You Get

• A Client object for managing indexes, keys, tasks, and instance-level settings
• An Index object for adding, updating, and deleting documents in bulk
• Full search support including filters, facets, highlighting, and hybrid semantic search
• Typed models via pydantic and camel-converter for clean Python-to-API field mapping
• Asynchronous task handling so you can track long-running indexing operations

Common Use Cases

• Adding instant, typo-tolerant search to a Django, Flask, or FastAPI application
• Bulk-indexing product catalogs, documentation, or content for a search-as-you-type UI
• Building faceted and filtered search experiences over structured datasets
• Powering hybrid keyword-plus-semantic search backed by a configured embedder

Under The Hood

Architecture — The client is organized around a small set of objects in the meilisearch package: Client (in client.py) is the entry point that manages indexes, keys, and instance-level operations, while Index (in index.py) handles document and search operations for a single index. All HTTP traffic funnels through _httprequests.py, which centralizes request building, auth headers, and error mapping; config.py holds connection settings and task.py models asynchronous task responses so long-running indexing calls can be tracked.

Tech Stack — Pure Python (3.10+) with only two runtime dependencies: requests for HTTP and camel-converter[pydantic] for mapping Meilisearch’s camelCase JSON onto snake_case pydantic models. The project uses setuptools for builds, uv for dependency management, and ships a py.typed marker for full type-hint support.

Code Quality — The repository has an extensive tests/ suite organized by concern (client, index, errors, settings, models) run under pytest with coverage and tox across Python versions, plus mypy and ruff for typing and linting. Errors are surfaced through a dedicated errors.py module rather than leaking raw HTTP exceptions, and the codebase is consistently typed.

API Design — The public surface is deliberately minimal and reads close to the Meilisearch documentation: instantiate a Client, call client.index(‘name’), then add_documents() and search(). Naming mirrors REST concepts, boilerplate is near-zero to get a first search running, and the docs plus in-repo examples make onboarding fast.