​
Synth AI OSS

Welcome to the Synth AI Open Source Software documentation. This section covers the core synth-ai Python package that provides a comprehensive framework for AI development, including language models, tracing, and environments.

​
What is Synth AI OSS?

Synth AI OSS is the open-source core of the Synth platform, providing:
  • Language Models: Unified interface for multiple LLM providers
  • Tracing: Comprehensive observability and debugging tools
  • Environments: Framework for creating AI-powered environments
  • Tools: Extensible tool system for agent-environment interaction

​
Quick Start

pip install synth-ai

from synth_ai.lm.core.main import LM

# Initialize a language model
lm = LM(model_name="gpt-4o-mini")

# Generate a response
response = lm.respond("Hello, world!")
print(response.raw_response)

​
Core Components

​
Language Models (synth_ai.lm)

Unified interface for interacting with various language model providers including OpenAI, Anthropic, and Google.

​
Tracing (synth_ai.tracing)

Comprehensive tracing system for debugging and analyzing AI applications, with persistent storage and session management.

​
Environments (synth_ai.environments)

Framework for creating stateful AI environments, including tools, registry, and lifecycle management.

​
Getting Started

  1. Installation: pip install synth-ai
  2. Basic Usage: See the API Reference for detailed documentation
  3. Examples: Check out the examples in the source code repository

​
Contributing

Synth AI OSS is open source! Contributions are welcome:
  • Report bugs and feature requests
  • Submit pull requests
  • Improve documentation
  • Add new environments and tools

​
License

This project is licensed under the MIT License.