AI Compare · Test kit
Edit an article without changing its facts
A 370-word local news draft with eleven figures, six dates, a quotation, an objection and two hedges. Cut it to about 300 words and check that every fact survives and nothing uncertain becomes certain.
Prices for United States, before sales tax Data release 2026.10.16-3, Oct 16, 2026 Rules 2.1.0
Run the test
- Download the file for the AI.
-
Copy the prompt.
Paste the whole file (the draft and the task under it) into the AI.
- Open the AI in another tab, in your own account, start a fresh chat, attach the file and paste the prompt.
- Note what you used: the plan, the visible model or mode, whether web search was on, how many attempts it took and anything that went wrong.
- Score the answer with the checks below. Run the same kit in each AI you are choosing between, or paste their answers into the blind review.
Scorecard
Critical checks are marked. One failed critical check matters more than a nice style.
- Keeps every figure: 7 to 2, 148 panels, 61,000 kWh, 40%, $186,000, $120,000, $66,000, $9,500, about seven years, about six weeks, 15 yearsCritical
- Keeps every date and time: Tuesday, October 6, 2026; 2023; March 2027; 2004; May 2026; Thursday, November 12, 2026 at 6:30 pmCritical
- Keeps Maribel Ortega's quotation word for wordCritical
- Keeps what is expected as expected, and the caveat that the savings could changeCritical
- Adds no fact, name or figure that is not in the draftCritical
- Keeps every name with its role: Maribel Ortega, Tom Reyes, Janet Cho
- Keeps Councillor Reyes's objection and the inspection that answered it
- Keeps who said what: the library's estimate, and 'the council said' on property taxes
- No more than 330 words
Compare answers blind
Paste the answer each AI gave, score them without seeing which AI wrote which, then reveal. You ran the tests yourself, so you may still recognise a style: hiding the names reduces that bias, it does not remove it. The answers stay in this tab and are never saved or sent anywhere; only the scores are kept, on this device, when you reveal.
Reference answer (open after you score)
Everything in the checklist is in the draft. Common failures: dropping 'about' from the payback period or the six weeks, writing that installation 'will begin' in March 2027, losing 'at current prices', turning the library's 40% estimate into a fact, giving the property-tax point to the library instead of the council, or cutting the objection to make the story one-sided.
Kit version 1.0.0, files frozen on Oct 13, 2026. Original fixture by What the tech, released under CC BY 4.0. Fictional people, companies and figures. A passing score shows how one answer did on this kit; it does not prove the tool is generally better.