fastapi-cache

Introduction

fastapi-cache is a tool to cache FastAPI endpoint and function results, with backends supporting Redis, Memcached, and Amazon DynamoDB.

Features

• Supports redis, memcache, dynamodb, and in-memory backends.
• Easy integration with FastAPI.
• Support for HTTP cache headers like ETag and Cache-Control, as well as conditional If-Match-None requests.
• Have an optional Cache Context available during function call, that can be acquired with get_cache_ctx function.

Requirements

• FastAPI
• redis when using RedisBackend.
• memcache when using MemcacheBackend.
• aiobotocore when using DynamoBackend.

Install

> pip install fastapi-cache2

or

> pip install "fastapi-cache2[redis]"

or

> pip install "fastapi-cache2[memcache]"

or

> pip install "fastapi-cache2[dynamodb]"

Usage

Quick Start

from collections.abc import AsyncIterator
from contextlib import asynccontextmanager

from fastapi import FastAPI
from starlette.requests import Request
from starlette.responses import Response

from fastapi_cache import FastAPICache
from fastapi_cache.backends.redis import RedisBackend
from fastapi_cache.decorator import cache

from redis import asyncio as aioredis


@asynccontextmanager
async def lifespan(_: FastAPI) -> AsyncIterator[None]:
    redis = aioredis.from_url("redis://localhost")
    FastAPICache.init(RedisBackend(redis), prefix="fastapi-cache")
    yield


app = FastAPI(lifespan=lifespan)


@cache()
async def get_cache():
    return 1


@app.get("/")
@cache(expire=60)
async def index():
    return dict(hello="world")

Initialization

First you must call FastAPICache.init during startup FastAPI startup; this is where you set global configuration.

Use the @cache decorator

If you want cache a FastAPI response transparently, you can use the @cache decorator between the router decorator and the view function.

You can also use the @cache decorator on regular functions to cache their result.

Currently, lock is available only with inmemory and redis backends (other backends will just ignore it).

Injected Request and Response dependencies

The cache decorator injects dependencies for the Request and Response objects, so that it can add cache control headers to the outgoing response, and return a 304 Not Modified response when the incoming request has a matching If-Non-Match header. This only happens if the decorated endpoint doesn't already list these dependencies already.

The keyword arguments for these extra dependencies are named __fastapi_cache_request and __fastapi_cache_response to minimize collisions. Use the injected_dependency_namespace argument to @cache to change the prefix used if those names would clash anyway.

Supported data types

When using the (default) JsonCoder, the cache can store any data type that FastAPI can convert to JSON, including Pydantic models and dataclasses, provided that your endpoint has a correct return type annotation. An annotation is not needed if the return type is a standard JSON-supported Python type such as a dictionary or a list.

E.g. for an endpoint that returns a Pydantic model named SomeModel, the return annotation is used to ensure that the cached result is converted back to the correct class:

from .models import SomeModel, create_some_model

@app.get("/foo")
@cache(expire=60)
async def foo() -> SomeModel:
    return create_some_model()

It is not sufficient to configure a response model in the route decorator; the cache needs to know what the method itself returns. If no return type decorator is given, the primitive JSON type is returned instead.

For broader type support, use the fastapi_cache.coder.PickleCoder or implement a custom coder (see below).

Custom coder

By default use JsonCoder, you can write custom coder to encode and decode cache result, just need inherit fastapi_cache.coder.Coder.

from typing import Any
import orjson
from fastapi.encoders import jsonable_encoder
from fastapi_cache import Coder

class ORJsonCoder(Coder):
    @classmethod
    def encode(cls, value: Any) -> bytes:
        return orjson.dumps(
            value,
            default=jsonable_encoder,
            option=orjson.OPT_NON_STR_KEYS | orjson.OPT_SERIALIZE_NUMPY,
        )

    @classmethod
    def decode(cls, value: bytes) -> Any:
        return orjson.loads(value)


@app.get("/")
@cache(expire=60, coder=ORJsonCoder)
async def index():
    return dict(hello="world")

Custom key builder

By default the default_key_builder builtin key builder is used; this creates a cache key from the function module and name, and the positional and keyword arguments converted to their repr() representations, encoded as a MD5 hash. You can provide your own by passing a key builder in to @cache(), or to FastAPICache.init() to apply globally.

For example, if you wanted to use the request method, URL and query string as a cache key instead of the function identifier you could use:

def request_key_builder(
    func,
    namespace: str = "",
    *,
    request: Request = None,
    response: Response = None,
    *args,
    **kwargs,
):
    return ":".join([
        namespace,
        request.method.lower(),
        request.url.path,
        repr(sorted(request.query_params.items()))
    ])


@app.get("/")
@cache(expire=60, key_builder=request_key_builder)
async def index():
    return dict(hello="world")

Cache Context

You can fetch current cache context with get_cache_ctx function. It returns a mutable mapping of type CacheCtx and has all caching parameters along with some variables set during cache initialisation.

This context can be used to override some caching variables if they depend on the function context:

from fastapi_cache import FastAPICache, get_cache_ctx

@app.get("/")
@cache(namespace="test")
async def cache_with_ctx(new_expire: int):
    ctx = get_cache_ctx()

    # override expiration with the provided value
    ctx["expire"] = new_expire
    return {"result": 42}

Using Locks with Cache

The with_lock parameter can be used to prevent multiple concurrent executions of computationally expensive functions. When enabled, if multiple requests try to access the same uncached endpoint simultaneously, only the first request will execute the function while others wait for the result to be cached. This is particularly useful for:

1. Preventing the "thundering herd" problem where multiple requests execute the same expensive operation simultaneously
2. Reducing database load by ensuring only one request computes a result
3. Improving response times for concurrent requests to uncached endpoints

Currently, lock functionality is available only with inmemory and redis backends.

Here's an example of using with_lock for a computationally expensive operation:

import asyncio
import time
from fastapi import FastAPI
from fastapi_cache.decorator import cache

app = FastAPI()

@app.get("/expensive-operation")
@cache(namespace="heavy_computation", with_lock=True, lock_timeout=30)
async def expensive_operation():
    # Simulate a heavy computation
    print("Starting expensive operation...")

    # This could be a database query, ML model inference, or any heavy computation
    await asyncio.sleep(5)  # Simulating 5 seconds of processing

    result = {"data": "Expensive computation result", "timestamp": time.time()}
    print("Expensive operation completed")

    return result

In this example:

• with_lock=True ensures that if multiple concurrent requests hit this endpoint, only one will execute the function
• lock_timeout=30 sets a maximum lock time of 30 seconds to prevent deadlocks
• After the first request completes, the result is cached and subsequent requests will get the cached result immediately

Without the lock, if 10 concurrent requests hit an uncached endpoint, all 10 would execute the expensive operation simultaneously. With the lock, only one request executes the operation while the other 9 wait for the result to be cached.

Backend notes

InMemoryBackend

The InMemoryBackend stores cache data in memory and only deletes when an expired key is accessed. This means that if you don't access a function after data has been cached, the data will not be removed automatically.

RedisBackend

When using the Redis backend, please make sure you pass in a redis client that does not decode responses (decode_responses must be False, which is the default). Cached data is stored as bytes (binary), decoding these in the Redis client would break caching.

Tests and coverage

coverage run -m pytest
coverage html
xdg-open htmlcov/index.html

License

This project is licensed under the Apache-2.0 License.