Skip to main content
Agent search asks AI assistants the questions your customers would ask, and records exactly how they answer. With ChatGPT selected, Rippit sends each question to an OpenAI model with its own web search turned on, and saves the answer, the sources it cited, and the searches it ran. For how research tables work overall, see How research tables work.

​
Setup

None. OpenAI models are built into Rippit. When you turn on agent search, pick one or more OpenAI models from the Search models list.

​
How it works

  1. Write the brief. It sets the topic, intent, and audience for the questions. “How do AI assistants recommend customer support QA tools to a mid-size SaaS company?”
  2. Rippit writes the questions. By default it generates 30 questions per input, grouped by theme. You can read, edit, add, or delete them before running.
  3. Each question goes to the model as-is, with web search on and no extra instructions from Rippit, so the answer reflects what the model says on its own.
  4. Every answer is saved as a row. Rippit repeats each question (twice by default) so you can see how consistent the answers are.
You can select up to three models in one table, across ChatGPT, Claude, and Gemini, to compare them on the same questions. Before a run, Rippit shows the total: questions × models × repeats.

​
What each row holds

One row per question, model, and repeat: Open a row’s trace to see everything behind an answer. “Not exposed” means the provider doesn’t return that detail. “None returned” means it does but there was nothing.

​
Good to know

  • These are API calls with search on, not a recording of the ChatGPT app. Answers can differ from what someone sees in the app.
  • A sample runs up to three questions on all selected models, once each.
  • Editing the brief asks you whether to regenerate the questions or keep the current ones. Questions you edited by hand are never replaced without asking.
  • Each new run is saved as its own study. The results show the latest, and older studies stay available in Agent search.

​
Good fits

  • Seeing whether AI assistants recommend your product, and what they say about competitors.
  • Checking whether answers about your product are accurate, and which sources they lean on.
  • Tracking how those answers change week to week on a schedule.