Can ChatGPT Cast a Prashna Kundali? What a General Chatbot Gets Wrong

Type "cast my prashna kundali" into ChatGPT and you will get a confident, well-formatted answer. It will name a lagna, describe the seventh house, quote a planet's degree and end with a verdict. The question this post answers is whether any of that is a Prashna Kundali at all — and how to check in thirty seconds.

What a Prashna Kundali Actually Is

Prashna Kundali — horary astrology — is a chart cast for the exact moment a question is put, at the seat of judgement: the place where the question is received and judged, traditionally the astrologer's own, not wherever the person asking happens to be. There is no birth data in it. Everything rests on three inputs: the instant of the question, the longitude and latitude of that seat, and an ephemeris that can turn those into planetary positions accurate to the minute of arc. From those positions come the lagna (rising sign), the house that carries the question, the Moon's condition and — in the KP method — the sub-lord of the relevant cusp, which is decided by arcs of a few minutes of longitude.

A chart that is off by fifteen minutes of time can change the lagna. A sub-lord can flip on less than a degree. Horary astrology is unforgiving about precision because precision is the entire method.

What a General-Purpose Chatbot Does Instead

A language model has no ephemeris. It does not know where Jupiter is right now unless the text it was trained on happened to say so, and it cannot compute planetary positions any more than it can compute your bank balance. When you ask ChatGPT to cast a prashna, one of four things happens:

1. It asks you for the positions. Some sessions will request the current planetary placements. If you can supply them from real software, the chatbot is only interpreting numbers you gave it — the casting was yours.
2. It approximates from memory. More often it states positions that are plausible for the year but wrong for the day, and always wrong for the lagna, which turns through all twelve signs every 24 hours and depends on the location the chart is cast for.
3. It invents a chart. Sometimes the positions are simply generated to look right. They are fluent, specific and false, and there is no way to tell from the text.
4. It gives a different chart next time. Ask the same question a minute later in a new session and the lagna, the house lords and the verdict will change, because nothing anchored the first answer to the sky.

None of this makes the model bad at astrology as a subject. It is excellent at explaining what a sub-lord is, why the Moon matters in horary, or what Prashna Marga says about a retrograde significator. Explanation is a language task. Casting is arithmetic on the sky, and a chatbot does not do arithmetic on the sky.

The Thirty-Second Test

Whatever tool you use — including ours — these five questions separate a cast chart from a written one:

• Did it record the exact time of the question and the place it was judged from? A prashna without both is a prose reading, not a chart.
• Can it show the planetary longitudes it used? Real casting produces degrees, minutes and a nakshatra pada for every planet. "Jupiter is strong in the ninth" without a degree is a guess.
• Does the lagna change with the minute? Ask two minutes apart; the lagna degree must move by about half a degree.
• Does it cite the rule it applied? "The eleventh cusp sub-lord signifies 2, 6 and 11" is checkable. "The energies are favourable" is not.
• Does it repeat itself? The same question, same moment, same place should give the same chart every time.

A general chatbot fails the first, second, third and fifth. It can pass the fourth only by quoting rules that were never applied to a real chart.

What a Purpose-Built Horary System Does Differently

ProxaAI's Prashna reading is not a prompt to a language model. The chart is cast first, by the Swiss Ephemeris, for the instant your question first arrives, at a fixed seat of judgement (Ujjain) — a follow-up on the same matter is judged on that same chart. The KP sub-lords of the relevant cusps are computed from those longitudes, not described from memory. Only then does the AI write. The specialists work from the cast chart and the computed facts, their notes are checked before the answer is written (a timing that no computed period supports is removed, for example), and the yes-or-no answer is combined from their verdicts by fixed weights. The methodology page walks through each stage.

The practical difference: two people asking the same question at the same minute get the same chart, and both can see it.

When ChatGPT Is the Right Tool

Use a general chatbot to learn — the difference between a horary and a natal chart, the meaning of a Moon in the twelfth, why some astrologers reject a chart with the lagna in the last degrees of a sign. Use it to draft the question you want to ask, which is half the craft: a prashna answers one clear question, asked once, with real intent.

Then have the chart cast by something that can see the sky. If you do not have a birth time, this is the one method that does not need it — the Prashna Kundali reading starts from the question alone, in English or Hindi, and shows you the chart it answered from. If you do know your birth details, a free kundli is the natal counterpart and needs nothing more than your date, time and place. The same problem, and a similar test, applies to birth charts: see Can ChatGPT make your kundli?