LangChain + LlamaIndex + AutoGen + CrewAI integration guide

README.md

@@ -1,3 +1,297 @@
----
-license: mit
----
+---
+license: mit
+tags:
+  - tutorial
+  - crazyrouter
+  - langchain
+  - llamaindex
+  - autogen
+  - ai-agents
+  - llm
+language:
+  - en
+  - zh
+---
+
+# 🔗 Crazyrouter + LangChain / LlamaIndex / AutoGen
+
+> Use 624+ AI models in your favorite framework — zero config changes.
+
+Since Crazyrouter is 100% OpenAI-compatible, it works out of the box with every major AI framework.
+
+---
+
+## LangChain
+
+### Installation
+
+```bash
+pip install langchain langchain-openai
+```
+
+### Basic Chat
+
+```python
+from langchain_openai import ChatOpenAI
+
+llm = ChatOpenAI(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="gpt-4o"
+)
+
+response = llm.invoke("Explain microservices in one paragraph")
+print(response.content)
+```
+
+### Switch Models on the Fly
+
+```python
+# Use Claude for analysis
+analyst = ChatOpenAI(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="claude-sonnet-4-20250514"
+)
+
+# Use DeepSeek for coding (cheaper)
+coder = ChatOpenAI(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="deepseek-chat"
+)
+
+# Use GPT-4o-mini for simple tasks (cheapest)
+helper = ChatOpenAI(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="gpt-4o-mini"
+)
+```
+
+### LangChain Chains
+
+```python
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_core.output_parsers import StrOutputParser
+
+prompt = ChatPromptTemplate.from_messages([
+    ("system", "You are a helpful coding assistant."),
+    ("user", "{question}")
+])
+
+chain = prompt | llm | StrOutputParser()
+
+result = chain.invoke({"question": "How do I read a CSV file in Python?"})
+print(result)
+```
+
+### RAG with Crazyrouter
+
+```python
+from langchain_openai import OpenAIEmbeddings, ChatOpenAI
+from langchain_community.vectorstores import FAISS
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_core.runnables import RunnablePassthrough
+
+# Embeddings (use a cheap model)
+embeddings = OpenAIEmbeddings(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="text-embedding-3-small"
+)
+
+# Chat model
+llm = ChatOpenAI(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="gpt-4o-mini"
+)
+
+# Split your documents
+splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
+docs = splitter.create_documents(["Your document text here..."])
+
+# Create vector store
+vectorstore = FAISS.from_documents(docs, embeddings)
+retriever = vectorstore.as_retriever()
+
+# RAG chain
+template = """Answer based on context:
+{context}
+
+Question: {question}"""
+
+prompt = ChatPromptTemplate.from_template(template)
+
+chain = (
+    {"context": retriever, "question": RunnablePassthrough()}
+    | prompt
+    | llm
+)
+
+result = chain.invoke("What does the document say?")
+print(result.content)
+```
+
+---
+
+## LlamaIndex
+
+### Installation
+
+```bash
+pip install llama-index llama-index-llms-openai-like
+```
+
+### Basic Usage
+
+```python
+from llama_index.llms.openai_like import OpenAILike
+
+llm = OpenAILike(
+    api_base="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="gpt-4o",
+    is_chat_model=True
+)
+
+response = llm.complete("What is retrieval augmented generation?")
+print(response)
+```
+
+### With OpenAI Class Directly
+
+```python
+from llama_index.llms.openai import OpenAI
+import os
+
+os.environ["OPENAI_API_KEY"] = "sk-your-crazyrouter-key"
+os.environ["OPENAI_API_BASE"] = "https://crazyrouter.com/v1"
+
+llm = OpenAI(model="gpt-4o-mini")
+response = llm.complete("Hello!")
+print(response)
+```
+
+---
+
+## AutoGen
+
+### Installation
+
+```bash
+pip install autogen-agentchat
+```
+
+### Multi-Agent Setup
+
+```python
+import autogen
+
+config_list = [
+    {
+        "model": "gpt-4o",
+        "base_url": "https://crazyrouter.com/v1",
+        "api_key": "sk-your-crazyrouter-key",
+    }
+]
+
+llm_config = {"config_list": config_list}
+
+# Create agents
+assistant = autogen.AssistantAgent(
+    name="assistant",
+    llm_config=llm_config,
+)
+
+user_proxy = autogen.UserProxyAgent(
+    name="user_proxy",
+    human_input_mode="NEVER",
+    max_consecutive_auto_reply=3,
+    code_execution_config={"work_dir": "coding"},
+)
+
+# Start conversation
+user_proxy.initiate_chat(
+    assistant,
+    message="Write a Python script that fetches the top 10 Hacker News stories."
+)
+```
+
+### Multi-Model Agents (Cost Optimization)
+
+```python
+# Expensive model for complex reasoning
+senior_config = [{"model": "gpt-4o", "base_url": "https://crazyrouter.com/v1", "api_key": "sk-your-key"}]
+
+# Cheap model for simple tasks
+junior_config = [{"model": "gpt-4o-mini", "base_url": "https://crazyrouter.com/v1", "api_key": "sk-your-key"}]
+
+senior = autogen.AssistantAgent("senior", llm_config={"config_list": senior_config})
+junior = autogen.AssistantAgent("junior", llm_config={"config_list": junior_config})
+```
+
+---
+
+## CrewAI
+
+```python
+from crewai import Agent, Task, Crew
+from langchain_openai import ChatOpenAI
+
+llm = ChatOpenAI(
+    base_url="https://crazyrouter.com/v1",
+    api_key="sk-your-crazyrouter-key",
+    model="gpt-4o"
+)
+
+researcher = Agent(
+    role="Researcher",
+    goal="Find accurate information",
+    backstory="You are an expert researcher.",
+    llm=llm
+)
+
+task = Task(
+    description="Research the latest trends in AI API gateways",
+    agent=researcher,
+    expected_output="A summary of trends"
+)
+
+crew = Crew(agents=[researcher], tasks=[task])
+result = crew.kickoff()
+print(result)
+```
+
+---
+
+## Environment Variables (Works Everywhere)
+
+Set these once and most frameworks auto-detect:
+
+```bash
+export OPENAI_API_KEY="sk-your-crazyrouter-key"
+export OPENAI_API_BASE="https://crazyrouter.com/v1"
+export OPENAI_BASE_URL="https://crazyrouter.com/v1"
+```
+
+---
+
+## Pro Tips
+
+1. **Use cheap models for agents that do simple tasks** — `gpt-4o-mini` or `deepseek-chat` for routing, summarizing, formatting
+2. **Use powerful models for reasoning** — `gpt-4o`, `claude-sonnet-4-20250514`, or `deepseek-reasoner` for complex analysis
+3. **Mix providers freely** — one agent uses Claude, another uses GPT, another uses Gemini. All through one API key
+4. **Monitor costs** — Crazyrouter dashboard shows per-model usage
+
+---
+
+## Links
+
+- 🌐 [Crazyrouter](https://crazyrouter.com)
+- 📖 [Getting Started Guide](https://huggingface.co/xujfcn/Crazyrouter-Getting-Started)
+- 🤖 [Live Demo](https://huggingface.co/spaces/xujfcn/Crazyrouter-Demo)
+- 💬 [Telegram](https://t.me/crazyrouter)
+- 🐦 [Twitter @metaviiii](https://twitter.com/metaviiii)