langchain-llm7

LangChain integration for the LLM7 Chat API.

This package provides a ChatLLM7 class that implements the LangChain SimpleChatModel interface, allowing seamless integration with the LangChain JS/TS ecosystem for both standard invocation and streaming responses.

Installation

npm install langchain-llm7 @langchain/core

or

yarn add langchain-llm7 @langchain/core

Note: @langchain/core is a peer dependency.

Usage

Here's how to use the ChatLLM7 model:

import { ChatLLM7 } from "langchain-llm7";
import { HumanMessage, SystemMessage } from "@langchain/core/messages";

// Initialize the model (defaults or provide specific options)
const chat = new ChatLLM7({
  // modelName: "gpt-4.1-nano", // Default
  // temperature: 0.8,
  // maxTokens: 150,
});

const messages = [
  new SystemMessage("You are a helpful assistant."),
  new HumanMessage("What is the capital of France?"),
];

// --- Basic Invocation (Non-streaming) ---
console.log("--- Testing invoke() ---");
try {
  const response = await chat.invoke(messages);
  console.log("Response:", response.content);
  // Example Output: Response: Paris
} catch (error) {
  console.error("Invoke Error:", error);
}

// --- Streaming ---
console.log("\n--- Testing stream() ---");
try {
  const stream = await chat.stream(messages);
  let fullResponse = "";
  process.stdout.write("Streamed Response: ");
  for await (const chunk of stream) {
    process.stdout.write(chunk.text); // Use .text for string content of the chunk
    fullResponse += chunk.text;       // Accumulate the text part
  }
  process.stdout.write("\n");
  console.log("(Full streamed content length:", fullResponse.length, ")");
  // Example Output: Streamed Response: Paris
  //                 (Full streamed content length: 5)
} catch (error) {
  console.error("Stream Error:", error);
}

Configuration

You can configure the ChatLLM7 model by passing parameters to its constructor:

| Parameter | Type | Default | Description | |--------------|--------------|------------------------------|-----------------------------------------------------------------------------| | baseUrl | string | "https://api.llm7.io/v1" | Base URL for the LLM7 API. | | modelName | string | "gpt-4.1-nano" | The specific LLM7 model to use. | | temperature| number | 1.0 | Sampling temperature (usually between 0 and 2). Higher values mean more randomness. | | maxTokens | number | undefined | Maximum number of tokens to generate in the completion. | | timeout | number | 120 | Request timeout in seconds. | | maxRetries | number | 3 | Maximum number of retries for failed API requests (network errors, 5xx, 429). | | stop | string[] | undefined | Optional list of sequences where the API should stop generating tokens. |

All standard BaseChatModelParams like callbacks, verbose, etc., are also accepted.

Development

1. Clone the repository: git clone https://github.com/chigwell/npm-langchain-llm7.git
2. Install dependencies: cd npm-langchain-llm7 && npm install
3. Build the package: npm run build
4. Run tests (using the example): npx ts-node test.ts

Contributing

Contributions are welcome! Please feel free to open an issue or submit a pull request on the GitHub repository.

License

This package is licensed under the Apache License 2.0.