Quickstart Guide#

This quickstart guide provides an introduction to the YandexGPT Python SDK, helping you get up and running quickly.

Obtaining Credentials#

To use the Yandex GPT API, you need to authenticate your requests using either an IAM token or an API key.

Remember to handle your tokens and keys securely and never expose them in your code or version control systems.

SDK Components#

  • YandexGPT Class: The YandexGPT class is the main interface to the Yandex GPT API. It provides methods to send both synchronous and asynchronous requests for text completions.

  • Configuration Managers: Configuration managers are used to set up and manage the necessary settings for the YandexGPT class, such as authentication and model details. For more information, see the Configuration Managers Module.

  • Threads: The SDK offers the YandexGPTThread class to manage conversation threads, allowing for maintaining the context of a conversation. For more details, refer to the YandexGPTThread Module.

Environmental Setup#

The /env directory contains .env.example files for each configuration manager. Copy the relevant example file to create your own .env file with the necessary credentials and settings.

Examples#

Here are quick examples to show how each component can be used:

Setting up a Configuration Manager

from yandex_gpt import YandexGPTConfigManagerForAPIKey

# Initialize the configuration manager with model type, catalog ID, and API key
config_manager = YandexGPTConfigManagerForAPIKey(
    model_type="yandexgpt",
    catalog_id="your_catalog_id",
    api_key="your_api_key"
)

Using the YandexGPT Class

from yandex_gpt import YandexGPT

# Initialize YandexGPT with a configuration manager
yandex_gpt = YandexGPT(config_manager=config_manager)

# Synchronous completion example
completion = yandex_gpt.get_sync_completion(messages=[{'role': 'user', 'text': 'Hello!'}])

print(completion)

Working with Threads

from yandex_gpt import YandexGPTThread

# Create a thread instance using the same configuration manager
thread = YandexGPTThread(config_manager=config_manager)

# Add a message to the thread and run it asynchronously
thread.add_message(role='user', text='Hello, Yandex GPT!')

# ... run and interact with the thread

Next Steps#

After going through these examples, you’re ready to dive deeper into the SDK. Explore the full documentation to learn more about advanced features and best practices for using the YandexGPT Python SDK.