Streaming Agent Execution in LangChain

Introduction

When we use agent.invoke(), we only see the final answer.

The agent may perform multiple steps internally:

1. Analyze the user request
2. Decide which tool to call
3. Execute the tool
4. Process the result
5. Generate a response

Normally, all of this happens behind the scenes.

Streaming allows us to observe the agent while it is working.

This is useful for:

• Debugging agents
• Building chat interfaces
• Showing progress to users
• Visualizing tool usage
• Understanding agent behavior

Why Streaming Matters

Consider the question:

Who is the richest person in Kochi?

Without streaming:

The user only sees the final response.

With streaming:

The application can display what the agent is doing in real time.

Streaming an Agent

Instead of using invoke(), use stream().

const stream = await agent.stream(
    {
        messages: [
            {
                role: 'human',
                content: 'Who is the richest person in Kochi?'
            }
        ]
    },
    {
        streamMode: 'values'
    }
);

The stream produces updates as the agent progresses through its workflow.

Complete Example

import { AIMessage, createAgent, tool } from 'langchain';
import * as z from 'zod';
import 'dotenv/config';

const searchTool = tool(
    async ({ input }) => {

        const url = new URL(
            `https://api.duckduckgo.com/?q=${encodeURIComponent(input)}&format=json`
        );

        const res = await fetch(url);

        if (!res.ok) {
            const body = await res.text();
            throw new Error(`Search API failed (${res.status}): ${body}`);
        }

        const data = await res.json();
        const results = data;

        return results ?? 'No results found';
    },
    {
        name: 'search',
        description: 'Search the web for information',
        schema: z.object({
            input: z.string().describe('The input to search for'),
        }),
    }
);

const agent = createAgent({
    model: 'gpt-5.4',
    tools: [searchTool],
});

const stream = await agent.stream({
        messages: [
            { role: 'human', content: 'Who is the richest person in Kochi?'}
        ],
    }, { streamMode : 'values'});

for await (const chunk of stream) {
    const latestMessage = chunk.messages.at(-1);

    if (latestMessage?.content) {
        console.log(`Agent: ${latestMessage.content}`);
    } else if (AIMessage.isInstance(latestMessage) && latestMessage.tool_calls?.length) {
        for (const call of latestMessage.tool_calls) {
            console.log(`Tool call: ${call.name}(${JSON.stringify(call.args)})`);
        }
    }
}