Deploy PinchTab

Deploy and Host Pinchtab on Railway

Pinchtab is a high-performance browser automation bridge built for AI agents. It exposes browser control through a fast HTTP API, supports multiple isolated browser instances, and includes a real-time dashboard for inspecting and managing automation sessions.

This template deploys Pinchtab with persistent storage and token-based authentication, giving AI agents a reusable browser runtime that can navigate websites, extract page structure, interact with elements, and run parallel browser workflows.

About Hosting Pinchtab

Pinchtab acts as a browser control layer between AI agents and Chromium.

Instead of embedding browser automation logic directly into every agent, applications can communicate with Pinchtab through its HTTP API. Agents can open pages, inspect interactive elements, click, fill forms, extract text, capture screenshots, and manage multiple browser instances.

Pinchtab also provides persistent profiles and configuration under /data, allowing browser-related state to survive service restarts and redeployments. The official Docker deployment uses the same persistent data path.

Included Architecture

                    ┌─────────────────┐
                    │   AI Agent      │
                    │ App / Workflow  │
                    └────────┬────────┘
                             │
                             │ HTTP API
                             ▼
                    ┌─────────────────┐
                    │    Pinchtab     │
                    │                 │
                    │ Browser API     │
                    │ Dashboard       │
                    │ Orchestration   │
                    └────────┬────────┘
                             │
                             ▼
                    ┌─────────────────┐
                    │    Chromium     │
                    │                 │
                    │ Tabs / Profiles │
                    │ Automation      │
                    └────────┬────────┘
                             │
                             ▼
                         Websites

Common Use Cases

• Give AI agents browser access
• Automate website navigation and interactions
• Extract structured or plain-text page content
• Fill forms and click interactive elements
• Run browser-based AI workflows
• Build web research agents
• Automate repetitive browser tasks
• Run multiple isolated browser profiles
• Monitor websites and extract changing data
• Integrate browser automation into APIs and workflows
• Connect coding agents and agent frameworks to a browser runtime

Pinchtab supports direct HTTP API access as well as an MCP interface with tools for navigation, interactions, screenshots, text extraction, network inspection, and browser management.

Pinchtab vs Traditional Browser Automation

Pinchtab is particularly useful when several applications or AI agents need reusable browser access without each application managing Chromium directly.

Browser Automation API

Pinchtab exposes its API on:

http://:9867

Typical automation operations include:

• Create browser profiles
• Start browser instances
• Open tabs
• Navigate to URLs
• Retrieve page snapshots
• Extract page text
• Find interactive elements
• Click elements
• Fill form fields
• Send keyboard input
• Capture screenshots
• Inspect network activity
• Manage multiple tabs and sessions

The official API examples use profile, instance, and tab resources to isolate parallel browser workflows.

Authentication

This template protects Pinchtab with an API token.

After deployment, open the Pinchtab service Variables page and copy:

PINCHTAB_TOKEN

API clients should send the token as:

Authorization: Bearer 

The dashboard also uses the token during login and exchanges it for a secure same-origin session cookie. Pinchtab strongly recommends using a token whenever the service is reachable beyond localhost.

Getting Started

1. Deploy the Pinchtab template.
2. Wait until the service is online.
3. Open the public Railway URL assigned to Pinchtab.
4. Copy PINCHTAB_TOKEN from the service Variables page.
5. Use the token to access the dashboard.
6. Create a browser profile.
7. Start a browser instance.
8. Open a tab.
9. Navigate to a website.
10. Use the API or dashboard to inspect and interact with the page.

Example API Flow

A typical Pinchtab workflow looks like:

Create Profile
      │
      ▼
Start Browser Instance
      │
      ▼
Open Tab
      │
      ▼
Navigate
      │
      ▼
Snapshot / Text
      │
      ▼
Click / Fill / Press

This makes Pinchtab suitable for agents that need to repeatedly observe a page, reason about available actions, and interact with the browser.

Multi-Instance Workflows

Pinchtab can run multiple isolated browser instances at the same time.

Each instance can use a separate profile, allowing workflows such as:

Pinchtab
   │
   ├── Profile A → Browser Instance A
   │
   ├── Profile B → Browser Instance B
   │
   └── Profile C → Browser Instance C

This is useful for:

• Parallel AI agents
• Multiple authenticated sessions
• Isolated customer workflows
• Concurrent scraping tasks
• Testing different browser identities

Pinchtab's instance API is specifically designed for starting and managing independent browser sessions.

MCP Support

Pinchtab also exposes MCP tooling for AI clients.

Its MCP integration includes tools for:

• Navigation
• Page snapshots
• Screenshots
• Clicking and typing
• Keyboard input
• Content extraction
• Tab management
• Waiting
• Network inspection
• Dialog handling

This allows compatible AI clients to interact with browser automation through MCP instead of calling the REST API manually.

Security Considerations

A remotely accessible browser automation API is highly privileged.

For production deployments:

• Always configure PINCHTAB_TOKEN.
• Do not expose the service without authentication.
• Keep the API token private.
• Use HTTPS for dashboard and API access.
• Limit which applications receive browser-control credentials.
• Restrict target domains when appropriate.
• Use isolated browser profiles for separate users or workloads.
• Review automation workflows before allowing them to access sensitive sites.

Pinchtab's security documentation warns that without a token, any process that can reach the service can invoke the browser API.

Dependencies for Pinchtab Hosting

• Pinchtab Server — browser automation API and orchestration layer
• Chromium — browser runtime used for automated sessions
• Persistent Volume — stores Pinchtab configuration and profiles
• Railway Public Networking — HTTPS access to the API and dashboard
• API Token — protects browser-control endpoints
• Shared Memory — important for browser stability under load

No PostgreSQL, Redis, or other external database is required for this template.

Deployment Dependencies

• Pinchtab GitHub: https://github.com/pinchtab/pinchtab
• Pinchtab Documentation: https://pinchtab.com/docs
• Pinchtab Docker Image: https://hub.docker.com/r/pinchtab/pinchtab
• Railway Networking: https://docs.railway.com/networking
• Railway Volumes: https://docs.railway.com/volumes

Why Deploy Pinchtab on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Pinchtab on Railway, you get a self-hosted browser automation runtime with persistent profiles, token-protected HTTP APIs, multi-instance browser orchestration, MCP support, and a real-time dashboard for AI agents and automation workflows.