OpenAI Operator logo

OpenAI Operator

other

by openai

AI agent that uses its own browser to navigate websites, complete tasks, and take actions on your behalf.

Key features

Web Browsing and Task Completion
Form Filling and Data Entry
Online Shopping
Booking and Reservations
Task Automation
Multi-tasking
Pricing

From $20/month

Best For

Power users wanting AI to handle web tasks like shopping, booking, research, and form filling without manual navigation

Verdict

Can complete real world tasks on actual websites, not just answer questions about them; it navigates, clicks, fills forms, and completes purchases

What it does

Web Browsing and Task Completion

Operator navigates real websites using its own browser, filling forms, clicking buttons, scrolling, and completing multi step tasks just as a human would. It handles login flows, shopping carts, booking systems, and complex web applications.

Learn more

Form Filling and Data Entry

Automatically fills out forms on any website by reading the form fields and entering the appropriate information. Handles shipping addresses, payment details (with your confirmation), registration forms, and application submissions.

Online Shopping

Browse products, compare prices, add items to carts, and proceed through checkout on e-commerce sites. Partner integrations with Instacart, DoorDash, and others provide an optimized experience for frequently used services.

Booking and Reservations

Make restaurant reservations (OpenTable), book travel (Priceline), schedule appointments, and handle other booking workflows across the web. The agent navigates the booking interface, selects options, and confirms on your behalf.

Task Automation

Automate repetitive web tasks that would normally require manual clicking and navigation. Research tasks, comparison shopping, data gathering from multiple sites, and multi step workflows that span several websites.

Multi-tasking

Run agent tasks in parallel across multiple conversations. Start a research task in one conversation, a shopping task in another, and a booking task in a third, all running simultaneously in the background.

Takeover Mode

For sensitive actions like entering payment information, confirming purchases, or submitting important forms, Operator pauses and hands control back to you. You complete the sensitive step manually, then the agent resumes.

Learn more

CUA Model via API

The Computer-Using Agent model is available through the OpenAI API for developers building custom agents. Send screenshots, receive action instructions, and build your own browser automation pipelines.

Learn more

Partner Integrations

Optimized workflows with DoorDash, Instacart, Uber, OpenTable, Priceline, StubHub, and other partners. These integrations provide faster, more reliable task completion than navigating partner sites manually through the browser.

Custom Instructions Per Site

Set specific instructions for how Operator should behave on particular websites. Define preferences for shopping sites, default options for booking platforms, and site specific behaviors.

Saved Prompts

Save frequently used task prompts for one click execution. Create saved prompts for recurring tasks like weekly grocery orders, regular restaurant reservations, or routine data gathering workflows.

Background Task Execution

Complex tasks that take 15 to 30 minutes run in the background while you continue using ChatGPT for other things. You receive a notification when the task is complete or when the agent needs your input.

Pricing

Best Value

Plus

$20/month

Access to agent mode within ChatGPT. Can browse the web and complete tasks autonomously. Included with ChatGPT Plus subscription.

  • Agent mode access in ChatGPT
  • Web browsing and task completion
  • Partner integrations
  • Takeover mode for sensitive actions
  • Saved prompts
  • Background task execution
  • Usage limits apply
Most accessible

Pro

$200/month

Full access with the highest limits. ChatGPT Pro subscription includes unlimited agent usage, priority access, and the most capable models.

  • All Plus features
  • Highest usage limits
  • Priority access
  • Most capable models
  • Unlimited agent usage
  • Advanced reasoning for complex tasks

API (CUA Model)

Custom

Build custom computer-using agents with the CUA model via the OpenAI API. Pay per token with no subscription required.

  • $3 per 1M input tokens
  • $12 per 1M output tokens
  • Full CUA model access
  • Send screenshots, receive actions
  • Build custom automation pipelines
  • No subscription required

Pros & Cons

Pros

  • Can complete real world tasks on actual websites, not just answer questions about them; it navigates, clicks, fills forms, and completes purchases
  • Partner integrations with DoorDash, Instacart, Uber, OpenTable, and others provide optimized, reliable task completion for common services
  • Three layer safety architecture prevents unauthorized actions: takeover mode for sensitive steps, data privacy protections, and adversarial defenses
  • CUA model available via API allows developers to build custom computer-using agents for specialized automation workflows
  • Background execution lets complex tasks (15 to 30 minutes) run while you continue working, with notifications when complete or when input is needed

Cons

  • No free tier; requires at minimum a ChatGPT Plus subscription ($20/month) to access agent mode
  • Not available in the EU/EEA or Switzerland due to regulatory restrictions, limiting access for European users
  • Reliability varies by website complexity; simple tasks succeed often, but complex multi step workflows on poorly designed sites can fail
  • Scored 38% on the OSWorld benchmark, showing significant limitations on complex desktop UI tasks and non standard web interfaces
  • Complex tasks can be slow, taking 15 to 30 minutes to complete workflows that a human might finish in 5 minutes with direct navigation
  • Financial transactions are restricted; the agent requires explicit user confirmation for purchases and cannot autonomously spend money
  • Prompt sensitivity affects success rates; slight changes in how you describe a task can lead to different outcomes or failures

How to get started

1

Subscribe to ChatGPT Plus or Pro

Operator requires at minimum a ChatGPT Plus subscription ($20/month). Pro ($200/month) gives the highest limits and most capable models. Sign up or upgrade at chatgpt.com.

2

Start an agent task

In ChatGPT, describe the task you want completed. For example: "Order groceries from Instacart: 2 lbs chicken breast, 1 bag of spinach, and a dozen eggs." Or: "Find and book a table for 2 at an Italian restaurant in San Francisco for Friday at 7pm." The agent will begin browsing and working on the task.

3

Watch and intervene when needed

The agent shows you its progress as it navigates websites. When it reaches a sensitive step (payment, login, important confirmation), it pauses and asks you to take over. Complete the sensitive action manually, then hand control back to the agent.

4

Save prompts for recurring tasks

If you find yourself requesting the same task regularly (weekly grocery orders, recurring reservations), save the prompt for one click execution in the future. Customize the prompt with your preferences and default options.

5

Try the CUA API for custom agents

Developers can access the Computer-Using Agent model via the OpenAI API. Send screenshots of a UI, receive action instructions (click at coordinates, type text, scroll), and build custom browser automation workflows.

Deep dive

Detailed guides with comparisons, tips, and visuals for each feature.

Get notified about updates

We'll email you when this tool's pricing or features change.

Last updated: 2026-02-21