Index

Index is the SOTA open-source browser agent for autonomously executing complex tasks on the web.

• Powered by reasoning LLMs with vision capabilities.
  • Gemini 2.5 Pro (really fast and accurate)
  • Claude 3.7 Sonnet with extended thinking (reliable and accurate)
  • OpenAI o4-mini (depending on the reasoning effort, provides good balance between speed, cost and accuracy)
  • Gemini 2.5 Flash (really fast, cheap, and good for less complex tasks)
• pip install lmnr-index and use it in your project
• index run to run the agent in the interactive CLI
• Index is also available as a serverless API.
• You can also try out Index via Chat UI.
• Supports advanced browser agent observability powered by open-source platform Laminar.

prompt: go to ycombinator.com. summarize first 3 companies in the W25 batch and make new spreadsheet in google sheets.

local_agent_spreadsheet_demo.mp4

Documentation

Check out full documentation here

Index API

The easiest way to use Index in production is via the serverless API. Index API manages remote browser sessions, agent infrastructure and browser observability. To get started, sign up and create project API key. Read the docs to learn more.

Install Laminar

pip install lmnr

Use Index via API

from lmnr import Laminar, AsyncLaminarClient
import asyncio
# you can also set LMNR_PROJECT_API_KEY environment variable

# Initialize tracing
Laminar.initialize(project_api_key="your_api_key")

# Initialize the client
client = AsyncLaminarClient(project_api_key="your_api_key")

async def main():

    response = await client.agent.run(
        prompt="Navigate to news.ycombinator.com, find a post about AI, and summarize it"
    )

    print(response.result)
    
if __name__ == "__main__":
    asyncio.run(main())

Local Quick Start

Install dependencies

pip install lmnr-index

# Install playwright
playwright install chromium

Setup model API keys

Setup your model API keys in .env file in your project root:

ANTHROPIC_API_KEY=
GEMINI_API_KEY=
OPENAI_API_KEY=

Run the agent with CLI

You can run Index via interactive CLI. It features:

• Browser state persistence between sessions
• Follow-up messages with support for "give human control" action
• Real-time streaming updates
• Beautiful terminal UI using Textual

You can run the agent with the following command. Remember to set API key for the selected model in the .env file.

index run

Output will look like this:

Loaded existing browser state
╭───────────────────── Interactive Mode ─────────────────────╮
│ Index Browser Agent Interactive Mode                       │
│ Type your message and press Enter. The agent will respond. │
│ Press Ctrl+C to exit.                                      │
╰────────────────────────────────────────────────────────────╯

Choose an LLM model:
1. Gemini 2.5 Flash
2. Claude 3.7 Sonnet
3. OpenAI o4-mini
Select model [1/2] (1): 3
Using OpenAI model: o4-mini
Loaded existing browser state

Your message: go to lmnr.ai, summarize pricing page

Agent is working...
Step 1: Opening lmnr.ai
Step 2: Opening Pricing page
Step 3: Scrolling for more pricing details
Step 4: Scrolling back up to view pricing tiers
Step 5: Provided concise summary of the three pricing tiers

Run the agent with code

import asyncio
from index import Agent, AnthropicProvider

async def main():

    llm = AnthropicProvider(
            model="claude-3-7-sonnet-20250219",
            enable_thinking=True, 
            thinking_token_budget=2048)
    # llm = OpenAIProvider(model="o4-mini") you can also use OpenAI models

    agent = Agent(llm=llm)

    output = await agent.run(
        prompt="Navigate to news.ycombinator.com, find a post about AI, and summarize it"
    )
    
    print(output.result)
    
if __name__ == "__main__":
    asyncio.run(main())

Stream the agent's output

async for chunk in agent.run_stream(
    prompt="Navigate to news.ycombinator.com, find a post about AI, and summarize it"
):
    print(chunk)

Enable browser agent observability

To trace Index agent's actions and record browser session you simply need to initialize Laminar tracing before running the agent.

from lmnr import Laminar

Laminar.initialize(project_api_key="your_api_key")

Then you will get full observability on the agent's actions synced with the browser session in the Laminar platform.

Run with remote CDP url

import asyncio
from index import Agent, AnthropicProvider, BrowserConfig

async def main():
    # Configure browser to connect to an existing Chrome DevTools Protocol endpoint
    browser_config = BrowserConfig(
        cdp_url="<cdp_url>"
    )
    
    llm = AnthropicProvider(model="claude-3-7-sonnet-20250219", enable_thinking=True, thinking_token_budget=2048)
    
    agent = Agent(llm=llm, browser_config=browser_config)
    
    output = await agent.run(
        prompt="Navigate to news.ycombinator.com and find the top story"
    )
    
    print(output.result)
    
if __name__ == "__main__":
    asyncio.run(main())

Customize browser window size

import asyncio
from index import Agent, AnthropicProvider, BrowserConfig

async def main():
    # Configure browser with custom viewport size
    browser_config = BrowserConfig(
        viewport_size={"width": 1200, "height": 900}
    )
    
    llm = AnthropicProvider(model="claude-3-7-sonnet-20250219")
    
    agent = Agent(llm=llm, browser_config=browser_config)
    
    output = await agent.run(
        "Navigate to a responsive website and capture how it looks in full HD resolution"
    )
    
    print(output.result)
    
if __name__ == "__main__":
    asyncio.run(main())

---

Made with ❤️ by the Laminar team