The Agent Agent Test

The Agent Agent test is a LLM qualitative Benchmark designed to assess a models ability at espionage. You can read more about the test here.

In order to use this test you need to provide your own OpenAi key. This key will not be stored but is used whilst on this page to communicate with OpenAI, you will be charged for use with each model. You can generate a specific key here (which you can revoke at anytime)