# Python AI SDK

## Docs

- [Basic text generation](https://pythonaisdk.mintlify.app/examples/basic-text.md): Your first ai_sdk call – generate a greeting with one function.
- [Vector search](https://pythonaisdk.mintlify.app/examples/embeddings-search.md): Embed a corpus and find the most similar sentence via cosine similarity.
- [Streaming responses](https://pythonaisdk.mintlify.app/examples/streaming.md): Display tokens as they arrive using stream_text.
- [Structured outputs](https://pythonaisdk.mintlify.app/examples/structured-output.md): Ask the model to return JSON validated against a Pydantic schema.
- [Tool-calling agents](https://pythonaisdk.mintlify.app/examples/tool-agent.md): Let the model call real Python functions to answer user queries.
- [Agent](https://pythonaisdk.mintlify.app/sdk/agent.md): Create AI agents with persistent state, system prompts, and tool-calling capabilities.
- [Concepts & architecture](https://pythonaisdk.mintlify.app/sdk/concepts.md): Understand the design goals and internal building blocks of ai_sdk.
- [cosine_similarity](https://pythonaisdk.mintlify.app/sdk/cosine_similarity.md): Calculate cosine similarity between two embedding vectors for semantic similarity analysis.
- [embed](https://pythonaisdk.mintlify.app/sdk/embed.md): Embed a single text value into a high-dimensional vector for semantic analysis.
- [embed_many](https://pythonaisdk.mintlify.app/sdk/embed_many.md): Embed multiple text values efficiently with automatic batching and retry logic.
- [generate_object](https://pythonaisdk.mintlify.app/sdk/generate_object.md): Parse model output directly into your own Pydantic models with automatic validation and retries.
- [generate_text](https://pythonaisdk.mintlify.app/sdk/generate_text.md): Generate text synchronously with a single call - returns a GenerateTextResult with rich metadata.
- [Introduction](https://pythonaisdk.mintlify.app/sdk/introduction.md): The Python AI SDK is a pure Python re-implementation of Vercel's popular AI SDK for TypeScript.
- [Providers](https://pythonaisdk.mintlify.app/sdk/providers.md): Factory functions that create ready-to-use model instances for different AI providers.
- [Anthropic Provider](https://pythonaisdk.mintlify.app/sdk/providers/anthropic.md): Claude models with OpenAI-compatible API interface for seamless integration.
- [OpenAI Provider](https://pythonaisdk.mintlify.app/sdk/providers/openai.md): Full OpenAI integration with GPT models, embeddings, function calling, and structured output.
- [stream_object](https://pythonaisdk.mintlify.app/sdk/stream_object.md): Stream structured output asynchronously with real-time validation and partial object updates.
- [stream_text](https://pythonaisdk.mintlify.app/sdk/stream_text.md): Stream text asynchronously for low-latency real-time generation with optional callbacks.
- [Tool Calling](https://pythonaisdk.mintlify.app/sdk/tool.md): Define and use tools with the AI SDK for function calling capabilities.
- [Data types](https://pythonaisdk.mintlify.app/sdk/types.md): Pydantic models used by the SDK – message hierarchy, token usage, tool calls and more.

## Optional

- [GitHub](https://github.com/python-ai-sdk/sdk)
- [PyPI](https://pypi.org/project/ai-sdk-python/)