An AI astrology reading, honestly explained
The AI does not guess your chart; an astronomy library computes it. The model’s job is the sentence-level work: turning your specific placements and aspects into a reading that answers your question. Here is exactly how that works, and where it stops.
The reading is optional · $3.99 one-time
Two halves: computation and interpretation
Every reading here has two separate parts. The first is arithmetic: given a date, time and place, an astronomy library computes where the Sun, Moon and planets were, to the arc-minute, and which sign and house each falls in under the Western tropical zodiac and Whole Sign houses. No AI is involved and nothing is estimated. The second part is language: a model is handed those positions and the aspects between them and asked to write six sections about them.
This matters because a chat assistant asked "what does my chart say" without exact positions will either ask you for them or invent them. The most common failure of AI astrology is not bad interpretation; it is a made-up chart. That failure cannot happen here because the model never computes anything.
What the model is told
The brief is specific. Each section is anchored to named placements: the headline to Sun, Moon and rising; how you think to Mercury and its tightest aspects; how you love to Venus and the seventh house; where you grow to Jupiter and the ninth; your edge to the tightest hard aspect touching a luminary or the angles; the undercurrent to Pluto, retrograde planets and the eighth and twelfth houses. It is told the orb of every aspect so it can weigh a 0.2° opposition over a 6° trine. It is told the length, two or three paragraphs, about 300 to 350 words, per section, and it is told to write in plain prose with no bullet points and no headings.
It is also told what not to do: no predictions of events or dates, no medical, legal or financial advice, no claims about people who are not in the chart. If you add a focus question, the brief includes it and the sections are written toward it.
What it cannot do
It cannot predict your future; nothing in astrology can, and we do not pretend otherwise. It cannot know anything about you that is not in the positions, so a wrong birth time gives a confidently wrong rising sign. It cannot hold a conversation the way a human astrologer does; you get one follow-up question, not a session. And it is entertainment and reflection, not a professional assessment of anything.
Why use AI for this at all
Cost and consistency. A consultation with a working astrologer runs from roughly eighty to several hundred dollars and is worth it for the dialogue. A written reading of your specific placements for $3.99, ready in a minute, is a different product, and the model does the sentence-level work at a price a human cannot. Every reading follows the same brief, so the sample shows exactly the structure you get. The chart itself, which is the hard part, is free either way.
What happens to your data
Your birth data is used to compute the chart and passed to the model to write the reading. The reading text is kept for 30 days under your payment reference so the link replays. There is no account, no email is collected, and the chart preview you share by link contains the chart only. A link shared from your finished reading includes the reading and is kept for 30 days.
Compute my chart free
Your chart is computed free, in full, before anything is offered. The reading is one payment, no account, secured by Whop.
Questions, answered
Which AI writes the reading?
Claude Sonnet 5, from Anthropic, working from a fixed brief and your computed placements. The brief, not the model, decides what each section is about.
Is this the same as asking ChatGPT about my chart?
No. A general assistant does not compute planetary positions; unless you paste an exact chart, it fills the gaps. Here the chart is computed first by an astronomy library and the model only interprets it.
Can the reading be wrong?
The positions cannot be, given correct birth data. The interpretation is astrology’s conventions applied to those positions, and like any reading it can weigh a placement more or less than you would. Treat it as a well-informed reflection, not a verdict.
Does the AI keep or learn from my data?
The reading text is stored for 30 days for replay. We do not build profiles, and we do not collect an email or create an account.