@axemere/gateway-langchain

LangChain BaseChatModel integration for the Axemere AI Gateway.

Installation

npm install @axemere/gateway-langchain @langchain/core

Usage

import { ChatAiGateway } from "@axemere/gateway-langchain";
import { HumanMessage } from "@langchain/core/messages";

const llm = new ChatAiGateway({ provider: "openai", model: "gpt-4o-mini" });
const res = await llm.invoke([new HumanMessage("Explain monads in one sentence.")]);
console.log(res.content);

Supported providers

openai, anthropic, mistral, google, xai, deepseek, groq, together, fireworks, perplexity, openrouter, cohere

Proxy-mode convenience re-exports

import { aiGatewayOpenAIClient, aiGatewayAnthropicClient } from "@axemere/gateway-langchain";

const openai = aiGatewayOpenAIClient();
const anthropic = aiGatewayAnthropicClient();

Configuration

Reads AXEMERE_GATEWAY_URL and AXEMERE_GATEWAY_TOKEN from the environment, or pass a config field to the constructor.

See @axemere/gateway for all configuration options.