On its own
A premium AI
with web search
- Waiting to start
The answer appears here.
The r0b test
Ask twice.
See the difference.
Your question goes to a premium AI on its own, and to a low-cost AI with r0b. Then both are challenged to say what each claim rests on, a check by code marks what holds, and you see what each answer cost.
How the test works
The same question.
Two ways to answer it.
01
Two answers
A premium AI answers on its own, with web search. A low-cost AI answers with r0b invoked first, as it would in your own AI app.
02
The challenge
Both get the same instruction: quote each factual claim and say what it rests on: a record given in this conversation, a web page read in it, or its own reasoning.
03
The check
Code checks every claim. A record must have come from r0b in this run and still open. A page must have been read in this run. A quote must be in the answer.
04
The verdict and the price
A blind judge says whether both reach the same decision. Each path is priced at OpenAI’s published rates, the challenge included.
Your question
- Checking the question
- Both AIs answer
- The challenge
- The check by code
- The verdict
With r0b invoked first
A low-cost AI with r0b
with r0b and web search
- Waiting to start
The answer appears here.
Checked: a record r0b reviewed, or a page that AI read in this runIts own reasoning: nothing to checkBroken: the record or page fails the check, or the quote is not in the answerUnmarked: not listed by that AI
The verdict
The judge sees both answers without knowing which is which, and says only whether they lead to the same decision. Declining to name a trade, a size or a timing is not a different decision on its own. Each AI chooses which claims to list; a page counts as checked when that AI read it in this run, under the same rule for both.
Bring r0b into your own AI
r0b. USD 497 a month,
or USD 4,970 a year.
14 days free. Add r0b to Claude, or to ChatGPT Pro, Business, Enterprise or Edu on the web, then use it in any question about the world outside your organisation.
Start 14 days free Buying for a team? hello@r0b.ai01
Grounded answers in the AI you already use: every claim labelled and checkable, as this test shows.
02
The world and its markets, rebuilt every night: situations, pressure and standing beliefs you can question, r0b’s market read and its tracked model portfolio (intelligence only, not financial advice).
03
Personal Intelligence Reports: some questions need more than a reply. Subscribers can commission a Personal Intelligence Report on any topic, at any time, with the brief agreed first.
The business case
Fewer premium seats.
r0b for the rest.
A premium seat buys more use of the same AI, not a different AI. In this test the low-cost AI with r0b reached the same decision for a fraction of the premium AI’s price, so fewer of your staff may need the premium allowance.
Seats moved20r0b for each one
r0b for themUSD 0a month
Seat savingUSD 0a month
How r0b is priced for a team. USD 497 for the first seat, 250 for the second, 125 for the third and 75 for the fourth, then USD 50 for every seat from the fifth. Teams are set up by hand: write to hello@r0b.ai.
Not a forecast. This is a business case from your numbers and the prices shown. Where the seat saving is smaller than r0b’s price, the case for r0b is what it gives you: grounded answers, the world and its markets, and Personal Intelligence Reports.
Keep the result
Start with r0b,
or take the result with you.
USD 497 a month, or USD 4,970 a year, with 14 days free.
Start 14 days free Buying for a team? hello@r0b.ai