AQ Challenge
Ten quick rounds, about two minutes. Test the practical AI judgment that actually matters at work — then share your tier and challenge a friend to the same rounds.
10 ROUNDS · 20 SECONDS EACH · 1 HINT
Ready to test your AI instincts?
Spot hallucinations, pick stronger prompts, catch privacy leaks and estimate token costs. Faster correct answers score more. Play with mouse, touch or keys 1–4 and H for a hint.
Daily AQ is the same 10 rounds for everyone, refreshed at 00:00 UTC.
What the AQ Challenge tests
Most AI quizzes ask trivia about model names. This game asks what you would actually do: which answer to distrust, which prompt to send, which data to keep out and which result to verify. Each game mixes six skill areas:
- Hallucination radarFind the fabricated citation, the invented API call or the summary whose numbers don't add up.
- Prompt craftChoose the prompt that names the audience, the facts, the format and what to do with missing data.
- Delegation judgmentDecide what AI can do alone, what needs your verification and what should not be delegated.
- Data privacySpot the secret, health detail or customer data that should never be pasted into a public chatbot.
- Token & cost senseEstimate tokens and API cost to the right order of magnitude — and pick the job that fits batch pricing.
- Verification & safetyRecognize prompt injection, misleading accuracy and agents that report green tests without proof.
How scoring works
A correct answer earns 7 points plus up to 3 for speed (within 5 seconds: +3, within 10: +2, within 20: +1). A round answered with the hint earns at most 6. Ten rounds give a score out of 100 and one of five tiers, from Fresh Starter to Frontier Operator. Daily AQ uses the same rounds for everyone each UTC day, so scores are comparable.
Privacy
Your answers stay in your browser. Share links contain only your score and the round set — no name or answers. We record a single completion event without parameters. The page shows Google ads: occasional full-screen ads between round blocks, an optional ad you can watch for a bonus hint in your next game, and one labelled display ad under the result. See our privacy policy.
Want the deeper version? The full AQ test maps five capability axes, and our practical AI tools help you plan costs, evaluations and pilots.