agents-langchain

1# LangChain - Build LLM Applications with Agents & RAG
2 
3The most popular framework for building LLM-powered applications.
4 
5## When to use LangChain
6 
7**Use LangChain when:**
8- Building agents with tool calling and reasoning (ReAct pattern)
9- Implementing RAG (retrieval-augmented generation) pipelines
10- Need to swap LLM providers easily (OpenAI, Anthropic, Google)
11- Creating chatbots with conversation memory
12- Rapid prototyping of LLM applications
13- Production deployments with LangSmith observability
14 
15**Metrics**:
16- **119,000+ GitHub stars**
17- **272,000+ repositories** use LangChain
18- **500+ integrations** (models, vector stores, tools)
19- **3,800+ contributors**
20 
21**Use alternatives instead**:
22- **LlamaIndex**: RAG-focused, better for document Q&A
23- **LangGraph**: Complex stateful workflows, more control
24- **Haystack**: Production search pipelines
25- **Semantic Kernel**: Microsoft ecosystem
26 
27## Quick start
28 
29### Installation
30 
31```bash
32# Core library (Python 3.10+)
33pip install -U langchain
34 
35# With OpenAI
36pip install langchain-openai
37 
38# With Anthropic
39pip install langchain-anthropic
40 
41# Common extras
42pip install langchain-community  # 500+ integrations
43pip install langchain-chroma     # Vector store
44```
45 
46### Basic LLM usage
47 
48```python
49from langchain_anthropic import ChatAnthropic
50 
51# Initialize model
52llm = ChatAnthropic(model="claude-sonnet-4-5-20250929")
53 
54# Simple completion
55response = llm.invoke("Explain quantum computing in 2 sentences")
56print(response.content)
57```
58…