Skip to main content
This integration method is maintained but no longer actively developed. For the best experience and latest features, use our new AI Gateway with unified API access to 100+ models.
1

Create an account + Generate an API Key

Log into helicone or create an account. Once you have an account, you can generate an API key.
2

Set required API keys as environment variables

# Required for all integrations
export HELICONE_API_KEY=<your Helicone API key>

# Required for specific LLM providers

export OPENAI_API_KEY=<your OpenAI API key>
export ANTHROPIC_API_KEY=<your Anthropic API key>
3

Initialize your LLM with Helicone

// Import the LangChain OpenAI package
import { ChatOpenAI } from "@langchain/openai";

// Initialize the OpenAI model with Helicone integration
const model = new ChatOpenAI({
  temperature: 0,
  modelName: "gpt-4o",
  configuration: {
    baseURL: "https://oai.helicone.ai/v1",
    defaultHeaders: {
      "Helicone-Auth": `Bearer ${process.env.HELICONE_API_KEY}`,
      "Helicone-Cache-Enabled": "true",
    },
  },
});
4

Add custom properties to your requests

You can add custom properties to your LangGraph agent invocations when calling invoke():
// Run the agent with custom Helicone properties
const result = await agent.invoke(
  { messages: [new HumanMessage("what is the current weather in sf")] },
  {
    options: {
      headers: {
        "Helicone-Prompt-Id": "weather-query",
        "Helicone-Session-Id": uuidv4(),
        "Helicone-Session-Name": "weather-session",
        "Helicone-Session-Path": "/weather",
      },
    },
  }
);
These properties will appear in your Helicone dashboard for filtering and analytics.

​
Complete Working Examples

​
OpenAI Example

// Import necessary dependencies
import { ChatOpenAI } from "@langchain/openai";
import { MemorySaver } from "@langchain/langgraph";
import { HumanMessage } from "@langchain/core/messages";
import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { TavilySearchResults } from "@langchain/community/tools/tavily_search";

// Define tools for the agent
const agentTools = [new TavilySearchResults({ maxResults: 3 })];

// Initialize the OpenAI model with Helicone integration
const agentModel = new ChatOpenAI({
  temperature: 0,
  modelName: "gpt-4o",
  configuration: {
    baseURL: "https://oai.helicone.ai/v1",
    defaultHeaders: {
      "Helicone-Auth": `Bearer ${process.env.HELICONE_API_KEY}`,
      "Helicone-Cache-Enabled": "true",
    },
  },
}).bindTools(agentTools);

// Initialize memory to persist state between graph runs
const agentCheckpointer = new MemorySaver();

// Create the agent
const agent = createReactAgent({
  llm: agentModel,
  tools: agentTools,
  checkpointer: agentCheckpointer,
});

// Run the agent with custom Helicone properties
const result = await agent.invoke(
  { messages: [new HumanMessage("what is the current weather in sf")] },
  {
    options: {
      headers: {
        "Helicone-Prompt-Id": "weather-query",
        "Helicone-Session-Id": uuidv4(),
        "Helicone-Session-Name": "weather-session",
        "Helicone-Session-Path": "/weather",
      },
    },
  }
);

​
Anthropic Example

// Import necessary dependencies
import { ChatAnthropic } from "@langchain/anthropic";
import { MemorySaver } from "@langchain/langgraph";
import { HumanMessage } from "@langchain/core/messages";
import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { TavilySearchResults } from "@langchain/community/tools/tavily_search";

// Define tools for the agent
const agentTools = [new TavilySearchResults({ maxResults: 3 })];

// Initialize the Anthropic model with Helicone integration
const agentModel = new ChatAnthropic({
  temperature: 0,
  modelName: "claude-3-opus-20240229",
  anthropicApiKey: process.env.ANTHROPIC_API_KEY,
  clientOptions: {
    baseURL: "https://anthropic.helicone.ai",
    defaultHeaders: {
      "Helicone-Auth": `Bearer ${process.env.HELICONE_API_KEY}`,
      "Helicone-Cache-Enabled": "true",
    },
  },
}).bindTools(agentTools);

// Initialize memory to persist state between graph runs
const agentCheckpointer = new MemorySaver();

// Create the agent
const agent = createReactAgent({
  llm: agentModel,
  tools: agentTools,
  checkpointer: agentCheckpointer,
});

// Run the agent with custom Helicone properties
const result = await agent.invoke(
  { messages: [new HumanMessage("what is the current weather in sf")] },
  {
    options: {
      headers: {
        "Helicone-Prompt-Id": "weather-query",
        "Helicone-Session-Id": uuidv4(),
        "Helicone-Session-Name": "weather-session",
        "Helicone-Session-Path": "/weather",
      },
    },
  }
);