Promptic : A simple yet powerful Python library for using large language models (LLMs).

Promptic

AI development assistant AI code assistant #Python #AI #LLM #litellm Standard Picks Open Source

Overview :

Promptic is a lightweight, decorator-based Python library that simplifies interactions with large language models (LLMs) through litellm. With promptic, you can easily create prompts, handle input parameters, and receive structured outputs from LLMs in just a few lines of code.

Target Users :

The target audience is developers and researchers who want to simplify their interactions with large language models (LLMs). Promptic provides a straightforward API and automates the handling of LLM responses, making it ideal for users who need to quickly implement and deploy LLM solutions.

Total Visits： 474.6M

Top Region： US(19.34%)

Website Views ： 52.7K

Use Cases

Create a function to query the President of the United States.

Use a Pydantic model to retrieve the query result for a country's capital.

Stream responses in real-time for generating poetry.

Features

Easily define prompts using decorators and function docstrings.

Automatically insert function parameters into prompts using {argument_name} placeholders in the docstring.

Support for Pydantic models to specify expected output structure, ensuring LLM responses conform to the defined schema.

Receive LLM responses in real-time by calling the decorated function with stream=True.

Simplify LLM interactions without needing to remember the shapes of OpenAPI response objects or other LLM-specific details.

Provide an easy-to-understand and reliable codebase leveraging the powerful capabilities of litellm, ensuring compatibility with a wide range of LLMs.

How to Use

Install the promptic library.

Import the llm decorator from the promptic library.

Define a function using the @llm decorator, and write the prompt in the function's docstring.

Use parameters in the function that will be automatically inserted into the prompt.

Call the function and pass the required parameters.

Process the structured output returned by the function.

(Optional) Use Pydantic models to define and validate the output structure.

(Optional) Set stream=True to stream LLM responses.