AsyncReducto Client

Overview

The AsyncReducto class is the asynchronous client for the Reducto API. It provides the same functionality as the synchronous Reducto client but with full async/await support for better performance in concurrent applications.

Constructor

from reducto import AsyncReducto

client = AsyncReducto(
    api_key="your-api-key",
    environment="production",
    timeout=60.0,
    max_retries=2,
)

Parameters

The AsyncReducto client accepts the same parameters as the synchronous Reducto client:

api_key

str

required

Your Reducto API key. If not provided, the client will automatically read from the REDUCTO_API_KEY environment variable.You can obtain your API key from the Reducto dashboard.

environment

Literal['production', 'eu', 'au']

default:"production"

The API environment to use. Available options:

production - US production environment (https://platform.reducto.ai)
eu - EU production environment (https://eu.platform.reducto.ai)
au - Australia production environment (https://au.platform.reducto.ai)

base_url

str | httpx.URL | None

Override the base URL for the API. If provided, this takes precedence over the environment parameter.You can also set this via the REDUCTO_BASE_URL environment variable.

timeout

float | Timeout | None

default:"3600.0"

Request timeout in seconds. Defaults to 1 hour (3600 seconds).For more granular control, pass an httpx.Timeout object:

import httpx
client = AsyncReducto(
    timeout=httpx.Timeout(60.0, read=5.0, write=10.0, connect=2.0)
)

max_retries

int

default:"2"

Maximum number of retries for failed requests. Defaults to 2.The client automatically retries:

Connection errors
408 Request Timeout
409 Conflict
429 Rate Limit
5xx Server errors

Set to 0 to disable retries.

default_headers

Mapping[str, str] | None

Additional headers to include with every request.

client = AsyncReducto(
    default_headers={"X-Custom-Header": "value"}
)

default_query

Mapping[str, object] | None

Default query parameters to include with every request.

http_client

httpx.AsyncClient | None

Custom httpx async client instance for advanced use cases. Use this to configure:

Custom proxies
Custom transports
Connection pooling

See the Using aiohttp section for an alternative HTTP client option.

Basic Usage

Simple Async Example

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto(
        # api_key defaults to REDUCTO_API_KEY environment variable
    )
    
    response = await client.parse.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    )
    
    print(response)

asyncio.run(main())

With Explicit API Key

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto(
        api_key="your-api-key",
    )
    
    response = await client.parse.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    )

asyncio.run(main())

With EU Environment

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto(
        environment="eu",
    )
    
    response = await client.parse.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    )

asyncio.run(main())

Client Methods

upload()

Upload a file directly to Reducto for processing. Returns a reducto:// URL that can be used in subsequent API calls.

import asyncio
from pathlib import Path
from reducto import AsyncReducto

async def main():
    client = AsyncReducto()
    
    # Upload a file
    upload_response = await client.upload(
        file=Path("document.pdf"),
        extension="pdf"  # Optional: specify file extension
    )
    
    print(upload_response.url)  # reducto://...
    
    # Use the uploaded file in a parse operation
    parse_response = await client.parse.run(
        input=upload_response.url,
    )

asyncio.run(main())

file

FileTypes | str | None

The file to upload. Can be:

A Path object
Raw bytes
A tuple of (filename, contents, media_type)

extension

str | None

Optional file extension hint (e.g., “pdf”, “docx”)

api_version()

Get the current API version.

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto()
    version = await client.api_version()
    print(version)

asyncio.run(main())

Available Resources

The AsyncReducto client provides access to the same resource namespaces as the synchronous client:

client.parse - Document parsing operations
client.split - Document splitting operations
client.extract - Data extraction operations
client.edit - Document editing operations
client.job - Job management operations
client.pipeline - Pipeline operations
client.webhook - Webhook configuration

All methods on these resources are async and must be awaited.

Using aiohttp

For improved concurrency performance, you can use aiohttp instead of the default httpx client:

Installation

pip install reductoai[aiohttp]

Usage

import asyncio
from reducto import AsyncReducto, DefaultAioHttpClient

async def main():
    async with AsyncReducto(
        http_client=DefaultAioHttpClient(),
    ) as client:
        response = await client.parse.run(
            input="https://pdfobject.com/pdf/sample.pdf",
        )
        print(response)

asyncio.run(main())

The aiohttp client provides better performance for applications making many concurrent requests.

Advanced Features

Context Manager

Use the async client as a context manager to ensure proper resource cleanup:

import asyncio
from reducto import AsyncReducto

async def main():
    async with AsyncReducto() as client:
        response = await client.parse.run(
            input="https://pdfobject.com/pdf/sample.pdf",
        )
        # HTTP client is automatically closed when exiting

asyncio.run(main())

Per-Request Configuration

Override client settings for individual requests:

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto()
    
    # Override timeout for a specific request
    response = await client.with_options(timeout=5.0).parse.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    )
    
    # Override max retries
    response = await client.with_options(max_retries=5).parse.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    )

asyncio.run(main())

Concurrent Requests

Process multiple documents concurrently for better performance:

import asyncio
from reducto import AsyncReducto

async def process_document(client, url):
    return await client.parse.run(input=url)

async def main():
    client = AsyncReducto()
    
    urls = [
        "https://example.com/doc1.pdf",
        "https://example.com/doc2.pdf",
        "https://example.com/doc3.pdf",
    ]
    
    # Process all documents concurrently
    results = await asyncio.gather(*[
        process_document(client, url) for url in urls
    ])
    
    for result in results:
        print(result)

asyncio.run(main())

Raw Response Access

Access raw HTTP response data including headers:

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto()
    
    response = await client.parse.with_raw_response.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    )
    
    print(response.headers.get('X-Request-Id'))
    parse_result = response.parse()  # Get the parsed response object

asyncio.run(main())

Streaming Response

Stream response data asynchronously for large payloads:

import asyncio
from reducto import AsyncReducto

async def main():
    client = AsyncReducto()
    
    async with client.parse.with_streaming_response.run(
        input="https://pdfobject.com/pdf/sample.pdf",
    ) as response:
        print(response.headers.get("Content-Type"))
        
        async for line in response.iter_lines():
            print(line)

asyncio.run(main())

Error Handling

The async client raises the same exceptions as the synchronous client:

import asyncio
import reducto
from reducto import AsyncReducto

async def main():
    client = AsyncReducto()
    
    try:
        response = await client.parse.run(
            input="https://pdfobject.com/pdf/sample.pdf",
        )
    except reducto.APIConnectionError as e:
        print("Network connection error")
        print(e.__cause__)
    except reducto.AuthenticationError as e:
        print("Invalid API key")
    except reducto.RateLimitError as e:
        print("Rate limit exceeded")
    except reducto.APIStatusError as e:
        print(f"API error: {e.status_code}")
        print(e.response)

asyncio.run(main())

Performance Tips

Use aiohttp for high concurrency: If you’re making many concurrent requests, the aiohttp client provides better performance than the default httpx client.
Reuse the client: Create one client instance and reuse it across requests to leverage connection pooling.
Use asyncio.gather(): Process multiple documents concurrently using asyncio.gather() for maximum throughput.
Adjust timeout settings: For long-running operations, increase the timeout to avoid premature failures.

Client

Resources

Types

Exceptions

Overview

Constructor

Parameters

Basic Usage

Simple Async Example

With Explicit API Key

With EU Environment

Client Methods

upload()

api_version()

Available Resources

Using aiohttp

Installation

Usage

Advanced Features

Context Manager

Per-Request Configuration

Concurrent Requests

Raw Response Access

Streaming Response

Error Handling

Performance Tips

See Also

Build docs developers (and LLMs) love

Client

Resources

Types

Exceptions

Documentation Index

​Overview

​Constructor

​Parameters

​Basic Usage

​Simple Async Example

​With Explicit API Key

​With EU Environment

​Client Methods

​upload()

​api_version()

​Available Resources

​Using aiohttp

​Installation

​Usage

​Advanced Features

​Context Manager

​Per-Request Configuration

​Concurrent Requests

​Raw Response Access

​Streaming Response

​Error Handling

​Performance Tips

​See Also

Build docs developers (and LLMs) love

Overview

Constructor

Parameters

Basic Usage

Simple Async Example

With Explicit API Key

With EU Environment

Client Methods

upload()

api_version()

Available Resources

Using aiohttp

Installation

Usage

Advanced Features

Context Manager

Per-Request Configuration

Concurrent Requests

Raw Response Access

Streaming Response

Error Handling

Performance Tips

See Also