Dify API Python

English | 中文

Dify API Python is a Python implementation for interacting with Dify's API. It provides an easy-to-use interface for making API calls, with support for both streaming and non-streaming responses.

Installation

Install the package using pip:

pip install dify-api-python

Usage

Importing the Client

from dify_api_python import DifyClient

Initializing the Client

You can initialize the DifyClient in several ways:

1. Using parameters:

   client = DifyClient(api_key='your_api_key', base_url='https://api.dify.ai/v1')

2. Using a default configuration file:

   client = DifyClient()

3. Using a custom configuration file:

   client = DifyClient(config_path='/path/to/your/config.ini')

The configuration file should be in INI format and include the following:

[DEFAULT]
API_KEY = your_api_key
BASE_URL = https://api.dify.ai/v1

Note: If you provide api_key or base_url as parameters, they will override any values in the configuration file.

Making API Calls

The DifyClient class acts as a wrapper for the DifySDK class. All methods available in the SDK can be called directly on the client object. Here are the available methods:

Chat Message

# Non-streaming
response = client.chat_message(
    query="Hello, how are you?",
    user="user123",
    inputs={},
    files=None,
    conversation_id=None,
    stream=False
)
print(response)

# Streaming
for chunk in client.chat_message(
    query="Tell me a story",
    user="user123",
    inputs={},
    files=None,
    conversation_id=None,
    stream=True
):
print(chunk)

Completion Message

# Non-streaming
response = client.completion_message(
    prompt="Once upon a time",
    user="user123",
    inputs={},
    stream=False
)
print(response)

# Streaming
for chunk in client.completion_message(
    prompt="Write a poem about",
    user="user123",
    inputs={},
    stream=True
):
    print(chunk)

Stop Chat Completion

result = client.stop_chat_completion(task_id, user="user123")

Get Conversation Messages

messages = client.get_conversation_messages(conversation_id, user="user123", first_id=None, limit=20)

Get Conversations

conversations = client.get_conversations(user="user123", last_id=None, limit=20, pinned=None)

Delete Conversation

result = client.delete_conversation(conversation_id, user="user123")

Rename Conversation

result = client.rename_conversation(conversation_id, name="New Name", user="user123", auto_generate=False)

Get Message Feedback

feedback = client.get_message_feedback(message_id, rating, user="user123")

Get Suggested Questions

questions = client.get_suggested_questions(message_id, user="user123")

Audio to Text

with open('audio.mp3', 'rb') as audio_file:
    result = client.audio_to_text(audio_file, user="user123")

Text to Audio

audio_content, content_type = client.text_to_audio(text="Hello, world!", message_id=None, user="user123")

Get App Parameters

parameters = client.get_app_parameters(user="user123")

Get App Meta

meta = client.get_app_meta(user="user123")

Upload File

with open('document.pdf', 'rb') as file:
    result = client.upload_file(file, user="user123")

Additional Client Methods

The DifyClient provides some additional methods:

# Combined chat completion (merges streaming response)
response = client.chat_completion_combined(
    query="Explain quantum computing",
    user="user123",
    inputs={}
)

# Reset a user's conversation
client.reset_conversation("user123")

# Get a user's conversation ID
conversation_id = client.get_conversation_id("user123")

Key Features

• Flexible client initialization (via parameters or config file)
• Support for both streaming and non-streaming API calls
• Method to merge streaming responses (chat_completion_combined)
• Easy-to-use interface for all Dify API endpoints
• Conversation management
• Audio conversion (text-to-speech and speech-to-text)
• File upload functionality

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the Apache License 2.0. See the LICENSE file for details.

Support

If you encounter any issues or have questions, please open an issue on the GitHub repository.