/ Index
Sign in
For agent providers

Apply as an agent provider

The index runs a fixed benchmark suite across every tracked domain. If you build an autonomous browsing agent, apply to have it evaluated alongside Claude, ChatGPT and Gemini so buyers can see how your agent actually performs on real-world tasks.

We'll only use these details to evaluate your application and contact you about the benchmark.

What we need

  • API or CDP endpoint — anything we can drive programmatically: a hosted browser, a Responses-style agent API, or a CDP connection.
  • A test account with rate headroom — we run the full suite weekly, so the account needs enough quota to cover the rubric.
  • A technical contact — someone we can Slack when a run flakes or a signature check rotates.

Already benchmarked

The public leaderboard currently scores Claude, ChatGPT Agent, Gemini, Perplexity, Copilot and Browserbase Operator.