> ## Documentation Index
> Fetch the complete documentation index at: https://docs.gulp.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Python SDK Introduction

> Get started with the Osmosis AI Python SDK

# Osmosis AI Python SDK

The Osmosis AI Python SDK provides tools for evaluating LLM outputs using natural language rubrics and reward functions. It supports both local deterministic evaluation and remote LLM-based semantic evaluation across multiple providers.

## What is Osmosis AI?

Osmosis AI helps you:

* **Evaluate LLM outputs** with natural language rubrics
* **Create reward functions** for reinforcement learning and scoring
* **Compare providers** across OpenAI, Anthropic, Gemini, xAI, and more
* **Batch process evaluations** with a built-in CLI tool

## Core Concepts

Osmosis AI provides two evaluation approaches:

**1. Local Reward Functions** - Fast, deterministic scoring (e.g., exact match, regex)
**2. Remote Rubric Evaluation** - LLM-powered semantic judgment with natural language criteria

See the [Quick Start](/python-sdk/quickstart) for code examples and hands-on tutorials.

### CLI Tool

Batch evaluate datasets from the command line:

```bash theme={null}
osmosis eval --rubric <rubric_id> --data <path_to_data>
```

See [CLI Quick Start](/python-sdk/cli-quickstart) for details.

## Key Features

<CardGroup cols={2}>
  <Card title="LLM-Based Rubrics" icon="brain">
    Natural language evaluation criteria with semantic understanding
  </Card>

  <Card title="Local Reward Functions" icon="bolt">
    Fast, deterministic functions for exact match and simple checks
  </Card>

  <Card title="Multi-Provider Support" icon="network-wired">
    OpenAI, Anthropic, Gemini, xAI, OpenRouter, and Cerebras
  </Card>

  <Card title="CLI Tool" icon="terminal">
    Batch evaluations with statistics and result tracking
  </Card>
</CardGroup>

## Use Cases

* **Quality Assurance** - Evaluate LLM responses before serving to users
* **Model Comparison** - Compare outputs across models and providers
* **Reinforcement Learning** - Create reward functions for training
* **A/B Testing** - Measure impact of prompt variations

## Next Steps

<CardGroup cols={2}>
  <Card title="Installation" icon="download" href="/python-sdk/installation">
    Install the SDK and set up API keys
  </Card>

  <Card title="Quick Start" icon="rocket" href="/python-sdk/quickstart">
    Your first evaluation in 5 minutes
  </Card>

  <Card title="CLI Quick Start" icon="terminal" href="/python-sdk/cli-quickstart">
    Batch evaluate datasets with the CLI
  </Card>

  <Card title="API Reference" icon="book" href="/python-sdk/api-reference">
    Complete API documentation
  </Card>
</CardGroup>