Agen & Multi-Agen

LangChain

langchain-ai/langchain

An open-source Python framework for building LLM-powered applications and agents through interoperable, chainable components and third-party integrations.

★ 143,6KBintang
⑂ 23,9KFork
451Isu terbuka
PythonBahasa
MITLisensi
Q@project.QualityScoreSkor editorial

Tangkapan layar proyek

Tangkapan layar LangChain Tangkapan layar LangChain

Ringkasan

LangChain is a framework for building agents and LLM-powered applications. It helps developers chain together interoperable components and third-party integrations to simplify AI application development while aiming to future-proof decisions as the underlying technology evolves. The framework provides a standard interface for models, embeddings, vector stores, and more, allowing developers to work at different levels of abstraction. It is part of a broader ecosystem that includes Deep Agents, LangGraph, and LangSmith.

Fitur utama

  • Real-time data augmentation via integrations with model providers, tools, vector stores, and retrievers
  • Model interoperability allowing models to be swapped during experimentation
  • Modular, component-based architecture for rapid prototyping
  • Production-ready features with support for monitoring, evaluation, and debugging
  • Flexible abstraction layers from high-level chains to low-level components
  • Seamless integration with the LangChain ecosystem including Deep Agents, LangGraph, and LangSmith

Persyaratan, instalasi, dan mulai cepat

The quickstart recommends using the `uv` package manager with the command: `uv add langchain`.

Penggunaan

A basic quickstart involves importing `init_chat_model` from `langchain.chat_models`, initializing a model (e.g., `init_chat_model("openai:gpt-5.5")`), and invoking it with a text prompt like `model.invoke("Hello, world!")`.

Kompatibilitas model dan kasus penggunaan

The framework supports model interoperability and integrates with various model providers. Repository topics indicate relevance to Anthropic, ChatGPT, Gemini, and OpenAI.

Catatan lisensi dan risiko

The repository is licensed under the MIT license.

Editorial verification 2026-08-09: repository URL, owner, description, license and repository statistics were reviewed. License metadata: MIT. README was fetched for the channel draft; re-check repository dependencies, releases and model terms before production use.

Rilis dan pemeliharaan

Not stated in the repository metadata.

Firecrawl

firecrawl/firecrawl

★ 161,1KTypeScript

RAGFlow

infiniflow/ragflow

★ 86,7KGo