Do different AI models really have distinguishable styles?

Yes, to a degree. Different training data, fine-tuning, and safety approaches give each model recognisable habits — Claude tends toward careful caveats and structure, ChatGPT toward upbeat, list-heavy answers, and Gemini toward concise, search-flavoured responses. The signals are probabilistic, not absolute, which is exactly what makes telling them apart a fun challenge.

How is this quiz scored?

Each round shows a response and asks which model produced it. You pick an answer and submit; the quiz immediately tells you whether you were right and explains the tells. Your running score is tracked, and at the end you get a percentage plus the rounds you missed so you can learn the patterns you struggle with.

Will identifying models this way always work in real life?

No — style overlaps heavily, models change with every update, and a single prompt can be answered in many ways. The quiz teaches general tendencies, not a reliable detector. Treat it as a way to build intuition about how the major models differ, not as a forensic tool for attributing any given text.

Do I need an account or API key to play?

No. The quiz runs entirely in your browser with a fixed set of questions and answers. Nothing is sent to any AI provider, no API key is required, and no personal data is collected. You can retake it as many times as you like.

"Which AI Said This?" Quiz: Identify the Model from its Response

Can you tell ChatGPT from Claude from Gemini?

The major AI assistants are trained differently, fine-tuned differently, and shaped by different safety philosophies — and it shows in how they write. This quiz presents real-style AI responses and challenges you to name the model behind each one. After every answer you get an explanation of the tell, so you finish with a sharper sense of how the leading models differ.

How the quiz works

Each round shows one AI-generated answer and asks which model produced it. Pick your choice and submit; the quiz instantly marks it right or wrong and explains the distinguishing signal — a Claude-style caveat, a ChatGPT-style upbeat list, a Gemini-style concise summary. Your score is tracked across all rounds, and at the end you see a percentage plus a list of the ones you missed.

Tips for spotting the model

Watch for a few recurring patterns. Claude often opens with careful framing, hedges with explicit caveats, and structures answers with headers or numbered reasoning. ChatGPT tends to be enthusiastic, list-heavy, and quick to offer a friendly summary or follow-up. Gemini frequently produces tighter, search-flavoured answers with a factual, slightly clipped tone. None of these are guarantees — styles overlap and change with every model update — so treat the quiz as intuition-building, not a forensic test, and have fun finding out how well you really know the AI you use every day.