llama-cloud

The official Python library for the llama-cloud API

MIT 91 个版本 Python >=3.9
Llama Cloud
安装
pip install llama-cloud
poetry add llama-cloud
pipenv install llama-cloud
conda install llama-cloud
描述

Llama Cloud Python SDK

PyPI version

The official Python SDK for LlamaParse - the enterprise platform for agentic OCR and document processing.

With this SDK, create powerful workflows across many features:

MCP Server

Use the Llama Cloud MCP Server to enable AI assistants to interact with this API, allowing them to explore endpoints, make test requests, and use documentation to help integrate this SDK into your application.

Add to Cursor Install in VS Code

Note: You may need to set environment variables in your MCP client.

Documentation

Installation

pip install llama_cloud

Quick Start

import os
from llama_cloud import LlamaCloud

client = LlamaCloud(
    api_key=os.environ.get("LLAMA_CLOUD_API_KEY"),  # This is the default and can be omitted
)

# Parse a document
job = client.parsing.create(
    tier="agentic",
    version="latest",
    file_id="your-file-id",
)

print(job.id)

File Uploads

from pathlib import Path
from llama_cloud import LlamaCloud

client = LlamaCloud()

# Upload using a Path
client.files.create(
    file=Path("/path/to/document.pdf"),
    purpose="parse",
)

# Or using bytes with a tuple of (filename, contents, media_type)
client.files.create(
    file=("document.txt", b"content", "text/plain"),
    purpose="parse",
)

Async Usage

import asyncio
from llama_cloud import AsyncLlamaCloud

client = AsyncLlamaCloud()


async def main():
    job = await client.parsing.create(
        tier="agentic",
        version="latest",
        file_id="your-file-id",
    )
    print(job.id)


asyncio.run(main())

MCP Server

Use the Llama Cloud MCP Server to enable AI assistants to interact with the API:

Add to Cursor Install in VS Code

Error Handling

When the API returns a non-success status code, an APIError subclass is raised:

import llama_cloud
from llama_cloud import LlamaCloud

client = LlamaCloud()

try:
    client.beta.indexes.list(
        project_id="my-project-id",
    )
except llama_cloud.APIConnectionError as e:
    print("The server could not be reached")
    print(e.__cause__)  # an underlying Exception, likely raised within httpx.
except llama_cloud.RateLimitError as e:
    print("A 429 status code was received; we should back off a bit.")
except llama_cloud.APIStatusError as e:
    print("Another non-200-range status code was received")
    print(e.status_code)
    print(e.response)
Status Code Error Type
400 BadRequestError
401 AuthenticationError
403 PermissionDeniedError
404 NotFoundError
422 UnprocessableEntityError
429 RateLimitError
>=500 InternalServerError
N/A APIConnectionError

Retries and Timeouts

The SDK automatically retries requests 2 times on connection errors, timeouts, rate limits, and 5xx errors. Requests timeout after 1 minute by default. Functions that combine multiple API calls (e.g. client.parsing.parse()) will have larger timeouts by default to account for the multiple requests and polling.

client = LlamaCloud(
    # default is 2
    max_retries=0,
)

# Or, configure per-request:
client.with_options(max_retries=5).beta.indexes.list(
    project_id="my-project-id",
)

Pagination

List methods support auto-pagination with for loops:

for run in client.extraction.runs.list(
    extraction_agent_id="agent-id",
    limit=20,
):
    print(run)

Or fetch one page at a time:

page = client.extraction.runs.list(extraction_agent_id="agent-id", limit=20)
for run in page.items:
    print(run)

while page.has_next_page():
    page = page.get_next_page()

Logging

Configure logging via the LLAMA_CLOUD_LOG environment variable or the log option:

client = LlamaCloud(
    log="debug",  # "debug" | "info" | "warn" | "error" | "off"
)

# More granular control:
client = LlamaCloud(
    timeout=httpx.Timeout(60.0, read=5.0, write=10.0, connect=2.0),
)

# Override per-request:
client.with_options(timeout=5.0).beta.indexes.list(
    project_id="my-project-id",
)

On timeout, an APITimeoutError is thrown.

Note that requests that time out are retried twice by default.

Advanced

Logging

We use the standard library logging module.

You can enable logging by setting the environment variable LLAMA_CLOUD_LOG to info.

$ export LLAMA_CLOUD_LOG=info

Or to debug for more verbose logging.

How to tell whether None means null or missing

In an API response, a field may be explicitly null, or missing entirely; in either case, its value is None in this library. You can differentiate the two cases with .model_fields_set:

if response.my_field is None:
  if 'my_field' not in response.model_fields_set:
    print('Got json like {}, without a "my_field" key present at all.')
  else:
    print('Got json like {"my_field": null}.')

Accessing raw response data (e.g. headers)

The "raw" Response object can be accessed by prefixing .with_raw_response. to any HTTP method call, e.g.,

from llama_cloud import LlamaCloud

client = LlamaCloud()
response = client.beta.indexes.with_raw_response.list(
    project_id="my-project-id",
)
print(response.headers.get('X-My-Header'))

index = response.parse()  # get the object that `beta.indexes.list()` would have returned
print(index.id)

These methods return an APIResponse object.

The async client returns an AsyncAPIResponse with the same structure, the only difference being awaitable methods for reading the response content.

.with_streaming_response

The above interface eagerly reads the full response body when you make the request, which may not always be what you want.

To stream the response body, use .with_streaming_response instead, which requires a context manager and only reads the response body once you call .read(), .text(), .json(), .iter_bytes(), .iter_text(), .iter_lines() or .parse(). In the async client, these are async methods.

with client.beta.indexes.with_streaming_response.list(
    project_id="my-project-id",
) as response:
    print(response.headers.get("X-My-Header"))

    for line in response.iter_lines():
        print(line)

The context manager is required so that the response will reliably be closed.

Making custom/undocumented requests

This library is typed for convenient access to the documented API.

If you need to access undocumented endpoints, params, or response properties, the library can still be used.

Undocumented endpoints

To make requests to undocumented endpoints, you can make requests using client.get, client.post, and other http verbs. Options on the client will be respected (such as retries) when making this request.

import httpx

response = client.post(
    "/foo",
    cast_to=httpx.Response,
    body={"my_param": True},
)

print(response.headers.get("x-foo"))

Undocumented request params

If you want to explicitly send an extra param, you can do so with the extra_query, extra_body, and extra_headers request options.

Undocumented response properties

To access undocumented response properties, you can access the extra fields like response.unknown_prop. You can also get all the extra fields on the Pydantic model as a dict with response.model_extra.

Configuring the HTTP client

You can directly override the httpx client to customize it for your use case, including:

import httpx
from llama_cloud import LlamaCloud, DefaultHttpxClient

client = LlamaCloud(
    # Or use the `LLAMA_CLOUD_BASE_URL` env var
    base_url="http://my.test.server.example.com:8083",
    http_client=DefaultHttpxClient(
        proxy="http://my.test.proxy.example.com",
        transport=httpx.HTTPTransport(local_address="0.0.0.0"),
    ),
)

You can also customize the client on a per-request basis by using with_options():

client.with_options(http_client=DefaultHttpxClient(...))

Managing HTTP resources

By default the library closes underlying HTTP connections whenever the client is garbage collected. You can manually close the client using the .close() method if desired, or with a context manager that closes when exiting.

from llama_cloud import LlamaCloud

with LlamaCloud() as client:
  # make requests here
  ...

# HTTP client is now closed

Versioning

This package generally follows SemVer conventions, though certain backwards-incompatible changes may be released as minor versions:

  1. Changes that only affect static types, without breaking runtime behavior.
  2. Changes to library internals which are technically public but not intended or documented for external use. (Please open a GitHub issue to let us know if you are relying on such internals.)
  3. Changes that we do not expect to impact the vast majority of users in practice.

We take backwards-compatibility seriously and work hard to ensure you can rely on a smooth upgrade experience.

We are keen for your feedback; please open an issue with questions, bugs, or suggestions.

Determining the installed version

If you've upgraded to the latest version but aren't seeing any new features you were expecting then your python environment is likely still using an older version.

You can determine the version that is being used at runtime with:

import llama_cloud
print(llama_cloud.__version__)

Requirements

  • Python 3.9+

Contributing

See CONTRIBUTING.md.

版本列表
2.9.1 2026-06-25
2.9.0 2026-06-09
2.8.0 2026-05-28
2.7.0 2026-05-20
2.6.0 2026-05-19
2.5.0 2026-05-14
2.4.1 2026-04-17
2.4.0 2026-04-16
2.3.0 2026-04-07
2.2.0 2026-04-03
2.1.0 2026-04-01
2.0.0 2026-03-31
1.6.0 2026-03-05
1.5.0 2026-03-03
1.4.1 2026-02-18
1.4.0 2026-02-12
1.3.0 2026-02-04
1.2.0 2026-01-30
1.1.0 2026-01-26
1.0.0 2026-01-21
1.0.0b7 2026-01-20
1.0.0b6 2026-01-19
1.0.0b5 2026-01-17
1.0.0b4 2026-01-16
1.0.0b3 2026-01-14
1.0.0b2 2026-01-07
1.0.0b1 2025-12-23
0.1.46 2026-01-21
0.1.45 2025-12-03
0.1.44 2025-11-04
0.1.43 2025-10-02
0.1.42 2025-09-16
0.1.41 2025-09-05
0.1.40 2025-08-29
0.1.39 2025-08-17
0.1.37 2025-08-04
0.1.36 2025-08-01
0.1.35 2025-07-28
0.1.34 2025-07-16
0.1.33 2025-07-08
0.1.32 2025-07-08
0.1.31 2025-07-07
0.1.30 2025-06-27
0.1.29 2025-06-25
0.1.28 2025-06-23
0.1.27 2025-06-20
0.1.26 2025-06-10
0.1.25 2025-06-10
0.1.24 2025-06-09
0.1.23 2025-05-28
0.1.22 2025-05-20
0.1.21 2025-05-08
0.1.20 2025-05-05
0.1.19 2025-04-25
0.1.18 2025-04-09
0.1.17 2025-03-28
0.1.16 2025-03-22
0.1.15 2025-03-18
0.1.14 2025-03-07
0.1.13 2025-02-19
0.1.12 2025-02-09
0.1.11 2025-01-27
0.1.10 2025-01-22
0.1.9 2025-01-15
0.1.8 2025-01-07
0.1.7 2024-12-23
0.1.6 2024-12-02
0.1.5 2024-11-12
0.1.4 2024-10-18
0.1.3 2024-10-18
0.1.2 2024-10-04
0.1.1 2024-10-03
0.1.0 2024-09-24
0.1.7a1 2024-12-21
0.0.17 2024-09-05
0.0.16 2024-09-05
0.0.15 2024-08-22
0.0.14 2024-08-21
0.0.13 2024-08-08
0.0.12 2024-08-06
0.0.11 2024-07-24
0.0.10 2024-07-23
0.0.9 2024-07-11
0.0.8 2024-07-09
0.0.7 2024-06-25
0.0.6 2024-06-21
0.0.5 2024-06-20
0.0.4 2024-06-20
0.0.3 2024-06-19
0.0.2 2024-06-19
0.0.1 2024-06-19