Table of Contents

Structured Outputs
- LangChain
- Instructor
- Marvin
- BAML
- Pydantic
- Magentic
- llm-structured-output
- LlamaIndex
- spacy-llm
- Outlines
- Guidance

Structured Outputs

LangChain

Website: https://www.langchain.com/
GitHub: https://github.com/langchain-ai/langchain
Features:
- Offers `.with_structured_output()` method
- PydanticOutputParser
- StructuredOutputParser

Instructor

GitHub: https://github.com/jxnl/instructor
PyPI: https://pypi.org/project/instructor/
Features:
- Built on Pydantic
- Facilitates generating structured output from LLMs

Marvin

Website: https://www.askmarvin.ai/
GitHub: https://github.com/PrefectHQ/marvin
Features:
- Tool for building reliable natural language interfaces
- Supports structured output capabilities

BAML

Website: https://www.baml.dev/
GitHub: https://github.com/baml-lang/baml
Features:
- Domain-specific language for writing and testing LLM functions

Pydantic

Website: https://docs.pydantic.dev/
GitHub: https://github.com/pydantic/pydantic
Features:
- Often used with other libraries to define and validate structured output schemas

Magentic

GitHub: https://github.com/jackmpcollins/magentic
PyPI: https://pypi.org/project/magentic/
Features:
- Works with Pydantic to generate complex structured outputs from LLMs

llm-structured-output

GitHub: https://github.com/KillianLucas/llm-structured-output
PyPI: https://pypi.org/project/llm-structured-output/
Features:
- Constrains LLM generation to structured output
- Includes function calling and JSON schema enforcement

LlamaIndex

Website: https://www.llamaindex.ai/
GitHub: https://github.com/jerryjliu/llama_index
Features:
- Supports both prompting and function calling for structured output

spacy-llm

GitHub: https://github.com/explosion/spacy-llm
PyPI: https://pypi.org/project/spacy-llm/
Features:
- Uses prompting for generating structured output

Outlines

GitHub: https://github.com/outlines-dev/outlines
PyPI: https://pypi.org/project/outlines/
Features:
- Employs constrained token sampling for structured output generation

Guidance

GitHub: https://github.com/microsoft/guidance
PyPI: https://pypi.org/project/guidance/
Features:
- Uses constrained token sampling for structured outputs