@langchain/classic

This package contains functionality from LangChain v0.x that has been moved out of the main langchain package as part of the v1.0 release. It exists to provide backward compatibility for existing applications while the core langchain package focuses on the essential building blocks for modern agent development.

When to use this package

Use @langchain/classic if you:

• Have existing code that uses legacy chains (e.g., LLMChain, ConversationalRetrievalQAChain, RetrievalQAChain)
• Use the indexing API
• Depend on functionality from @langchain/community that was previously re-exported from langchain
• Are maintaining an existing application and not yet ready to migrate to the new createAgent API

When NOT to use this package

For new projects, use langchain v1.0 instead. The new APIs provide:

• createAgent: A cleaner, more powerful way to build agents with middleware support
• Better performance: Optimized for modern agent workflows
• Focused API surface: Less complexity, easier to learn
• Active development: New features and improvements will focus on v1.0 APIs

See the LangChain v1.0 release notes for more information.

Installation

npm install @langchain/classic

This package requires @langchain/core as a peer dependency:

npm install @langchain/core

What's included

Legacy Chains

All chain implementations from v0.x, including:

• LLMChain - Basic chain for calling an LLM with a prompt template
• ConversationalRetrievalQAChain - Chain for conversational question-answering over documents
• RetrievalQAChain - Chain for question-answering over documents without conversation memory
• StuffDocumentsChain - Chain for stuffing documents into a prompt
• MapReduceDocumentsChain - Chain for map-reduce operations over documents
• RefineDocumentsChain - Chain for iterative refinement over documents
• And many more...

Indexing API

The RecordManager and related indexing functionality for managing document updates in vector stores.

Community Integrations

Re-exports from @langchain/community that were previously available in the main langchain package.

Other Deprecated Functionality

Various utilities and abstractions that have been replaced by better alternatives in v1.0.

Migration

From langchain v0.x to @langchain/classic

If you're upgrading to langchain v1.0 but want to keep using legacy functionality:

1. Install @langchain/classic:

   npm install @langchain/classic

2. Update your imports:

   // Before (v0.x)
   import { LLMChain } from "langchain/chains";
   import { ConversationalRetrievalQAChain } from "langchain/chains";
   
   // After (v1.0)
   import { LLMChain } from "@langchain/classic/chains";
   import { ConversationalRetrievalQAChain } from "@langchain/classic/chains";

   Or if you imported from the root:

   // Before (v0.x)
   import { LLMChain } from "langchain";
   
   // After (v1.0)
   import { LLMChain } from "@langchain/classic";

From @langchain/classic to langchain v1.0

For new development, we recommend using createAgent instead of legacy chains.

Example migration from LLMChain:

// Before (using LLMChain)
import { LLMChain } from "@langchain/classic/chains";
import { ChatOpenAI } from "@langchain/openai";
import { PromptTemplate } from "@langchain/core/prompts";

const model = new ChatOpenAI({ model: "gpt-4" });
const prompt = PromptTemplate.fromTemplate(
  "What is a good name for a company that makes {product}?"
);
const chain = new LLMChain({ llm: model, prompt });
const result = await chain.call({ product: "colorful socks" });

// After (using createAgent)
import { createAgent } from "langchain";

const agent = createAgent({
  model: "openai:gpt-4",
  systemPrompt: "You are a creative assistant that helps name companies.",
});

const result = await agent.invoke({
  messages: [
    {
      role: "user",
      content: "What is a good name for a company that makes colorful socks?",
    },
  ],
});

For more complex migrations, see the migration guide.

Support and Maintenance

@langchain/classic will receive:

• Bug fixes: Critical bugs will be fixed
• Security updates: Security vulnerabilities will be patched
• No new features: New functionality will focus on langchain v1.0 APIs

This package is in maintenance mode. For new features and active development, use langchain v1.0.

Examples

Using a legacy chain

import { LLMChain } from "@langchain/classic/chains";
import { ChatOpenAI } from "@langchain/openai";
import { PromptTemplate } from "@langchain/core/prompts";

const model = new ChatOpenAI({ model: "gpt-4" });

const prompt = PromptTemplate.fromTemplate(
  "Tell me a {adjective} joke about {content}."
);

const chain = new LLMChain({ llm: model, prompt });

const result = await chain.call({
  adjective: "funny",
  content: "chickens",
});

console.log(result.text);

Using ConversationalRetrievalQAChain

import { ConversationalRetrievalQAChain } from "@langchain/classic/chains";
import { ChatOpenAI } from "@langchain/openai";
import { OpenAIEmbeddings } from "@langchain/openai";
import { MemoryVectorStore } from "langchain/vectorstores/memory";

// Create vector store with documents
const vectorStore = await MemoryVectorStore.fromTexts(
  ["Document 1 text...", "Document 2 text..."],
  [{ id: 1 }, { id: 2 }],
  new OpenAIEmbeddings()
);

// Create chain
const model = new ChatOpenAI({ model: "gpt-4" });
const chain = ConversationalRetrievalQAChain.fromLLM(
  model,
  vectorStore.asRetriever()
);

// Use chain
const result = await chain.call({
  question: "What is in the documents?",
  chat_history: [],
});

console.log(result.text);

Resources

• LangChain v1.0 Release Notes
• Migration Guide
• LangChain v1.0 Documentation
• GitHub Repository

Support

For bug reports and issues, please open an issue on GitHub.

For questions and discussions, join our Discord community.

License

This package is licensed under the MIT License. See the LICENSE file for details.