PinchTab: High-Performance Browser Automation for AI Agents

Introduction

PinchTab is a high-performance browser automation bridge and multi-instance orchestrator, specifically designed to empower AI agents with direct control over Chrome. It operates as a standalone HTTP server, providing a token-efficient way for agents to navigate pages, extract structured content, and interact with the DOM without relying on raw HTML or images. Built in Go, PinchTab is a small, self-contained binary that emphasizes a local-first security posture, making it ideal for secure, single-user control while also supporting advanced remote and distributed deployments.

It enables a wide range of applications, from headless and headed navigation for data extraction and testing, to managing multiple isolated Chrome instances for complex distributed automation tasks.

Installation

Getting started with PinchTab is straightforward. Choose the method that best suits your environment:

macOS / Linux:

curl -fsSL https://pinchtab.com/install.sh | bash

Homebrew (macOS / Linux):

brew install pinchtab/tap/pinchtab

npm:

npm install -g pinchtab

Docker:

For containerized environments, you can run PinchTab with Docker:

docker run -d \
  --name pinchtab \
  -p 127.0.0.1:9867:9867 \
  -v pinchtab-data:/data \
  --shm-size=2g \
  pinchtab/pinchtab

For daily local use, it is recommended to install the daemon once to keep PinchTab running in the background:

pinchtab daemon install
pinchtab daemon

Examples

PinchTab offers both a powerful CLI and a flexible HTTP API for controlling browser instances.

CLI Usage

First useful command, auto-starts the local server if needed:

pinchtab nav https://pinchtab.com --snap

Recommended for daily local use, install the daemon once:

pinchtab daemon install
pinchtab daemon

Control the browser from another terminal:

# Navigate, starts the server automatically if needed
pinchtab nav https://pinchtab.com

# Get page structure
pinchtab snap -i -c

# Click an element
pinchtab click e5

# Extract text
pinchtab text

HTTP API Usage

You can also interact directly with the HTTP API:

# Create a profile first (returns profile id)
PROF=$(curl -s -X POST http://localhost:9867/profiles \
  -H "Content-Type: application/json" \
  -d '{"name":"work"}' | jq -r '.id')

# Start an instance for that profile (returns instance id)
INST=$(curl -s -X POST http://localhost:9867/instances/start \
  -H "Content-Type: application/json" \
  -d "{\"profileId\":\"$PROF\",\"mode\":\"headless\"}" | jq -r '.id')

# Open a tab in that instance
TAB=$(curl -s -X POST http://localhost:9867/instances/$INST/tabs/open \
  -H "Content-Type: application/json" \
  -d '{"url":"https://pinchtab.com"}' | jq -r '.tabId')

# Get snapshot
curl "http://localhost:9867/tabs/$TAB/snapshot?filter=interactive"

# Click element
curl -X POST "http://localhost:9867/tabs/$TAB/action" \
  -H "Content-Type: application/json" \
  -d '{"kind":"click","ref":"e5"}'

Why Use PinchTab?

PinchTab stands out for its robust feature set and performance optimizations, making it an excellent choice for browser automation, especially for AI agents:

Token Performance: Highly efficient, extracting text with approximately 800 tokens per page, significantly cheaper than screenshot-based methods.
Headless and Headed Modes: Supports both invisible headless Chrome for background tasks and visible headed Chrome for interactive workflows.
Persistent Profiles: Manage browser state, including cookies and history, allowing agents to maintain logged-in sessions across restarts.
Advanced CDP Control: Offers fine-grained control over Chrome DevTools Protocol.
Multi-instance Orchestration: Run multiple parallel Chrome processes with isolated profiles, ideal for distributed tasks and testing.
Self-contained Binary: A small, ~15MB Go binary with no external dependencies, simplifying deployment.
Accessibility-first: Uses stable element references instead of fragile coordinates for reliable interaction.
ARM64-optimized: First-class support for ARM64 architectures, including Raspberry Pi.

Benchmarking against other agent-browser solutions shows PinchTab significantly reduces token costs and requests, with cost savings widening for longer workloads.

PinchTab is also committed to privacy, operating as a fully open-source, local-first tool with no telemetry or analytics.