Python client for the Langflow API

This package provides an easy way to use the Langflow API to run flows from Python applications.

• Package: langflow-client
• Import: import langflow_client
• License: Apache-2.0

Installation

pip install langflow-client

Prerequisites

• Base URL of your Langflow instance
• API key if authentication is enabled on your instance

Usage

Initialization

from langflow_client import LangflowClient

client = LangflowClient({
    "base_url": "http://localhost:7860",
    "api_key": "sk-...",  # optional
})

Running a flow

flow = client.flow("<flow-id>")
response = flow.run("Hello, world!")
print(response.outputs)
print(response.chat_output_text())

You can pass tweaks and IO options:

from langflow_client.consts import InputTypes, OutputTypes

response = flow.run(
    "Hello",
    input_type=InputTypes.CHAT,
    output_type=OutputTypes.CHAT,
    session_id="optional-session-id",
    tweaks={"ChatInput-abc": {"temperature": 0.2}},
)

Streaming

for event in flow.stream("Hello"):
    print(event)

Events are dicts with event keys such as add_message, token, and end.

File upload

• v1 flow-scoped image upload:

with open("image.jpg", "rb") as f:
    upload = flow.upload_file(f, filename="image.jpg", content_type="image/jpeg")
    print(upload.file_path)

• v2 user-scoped file upload and listing:

with open("document.pdf", "rb") as f:
    uploaded = client.files.upload(f, filename="document.pdf", content_type="application/pdf")

files = client.files.list()

License

Apache License 2.0. See LICENSE