gptfree.net
WebGames

WebGames

Test AI web agents

Challenges for web AI agents

AI modelDevelopment and Tools AI TestingWeb BrowsingChallengesPassword VerificationTimer
Visit official site →

Highlights

Challenges for web browsing AI agents
Unique password on task completion
Time based, visual, and logic challenges
Free to use, aimed at researchers

What you can do with it

  • Test an agent's browsing ability
  • Study AI versus human performance
  • Benchmark progress on web tasks

How to use it

  1. 1 Open the WebGames site and pick a challenge.
  2. 2 Read the instructions and goal.
  3. 3 Complete the task by clicking, dragging, or typing.
  4. 4 Collect the unique password on success.
  5. 5 Note the completion time for comparison.

Who it is for

AI developers testing agentsResearchers studying human and AI differencesCurious users exploring agent limits

Pros

  • Clear pass proof through unique passwords
  • Varied challenge types keep tests broad
  • Timing gives a simple performance measure

Limitations

  • Aimed at developers rather than casual players
  • Free status can change without notice
  • Difficulty varies widely between tasks
  • Manual attempts take real effort without an agent

What WebGames is

WebGames is a set of challenges built to test general purpose web browsing AI agents. The tasks are easy for humans but hard for agents, and each completed challenge returns a unique password as proof. Challenge types include time based tasks, visual recognition, and logical reasoning, with interactions such as clicking, dragging, and typing. A timer records completion time, and it was created by convergence.ai to study how agents handle the open web. The platform is currently free and aimed at AI researchers and developers who want to measure agent progress.

What you can do with it

- Run a browsing agent against a challenge

- Verify success through the returned password

- Compare performance across challenge types

- Record completion time for each task

- Explore where current agents still struggle

The password system keeps scoring simple, and a timer makes repeat attempts easy to compare.

Who it is for

- AI developers testing their agents

- Researchers studying human and AI differences

- Curious users exploring agent limits

- Teams benchmarking progress over time

Educators use it as a demo of current AI limits in class, and developers get the most value, though anyone can try a task by hand.

What to watch out for

The platform is aimed at developers rather than casual players, so expect rough edges in presentation. Free status can change without notice. Difficulty varies widely between tasks, so a single result says little about overall capability. Manual attempts take real effort, and building or configuring an agent takes more. Tasks may involve real websites, so respect their terms while testing. Features and interface details may change between versions, so review the latest notes on the page before you start.

FAQ

What is WebGames?

A collection of challenges that test general purpose web browsing AI agents.

How do I know a task is solved?

Each completed challenge provides a unique password as proof.

Is it free?

The listing says the platform is currently free.

Why it is free

What the official page says: 免费

Free limits, trial length and features are decided by the official site and can change at any time.

Product preview

WebGames product screenshot

Screenshot archived from the listing page, for preview only.

WebGames

A collection of challenges designed to test general-purpose web browsing AI agents.

Visit official site →

Related free tools