Real charts. AI prose. That's the rule, and we mean both halves of it.
AstroGuide uses large language models to turn structured astrological data into readable prose. We want to tell you exactly what that means — and what it doesn't — so you can decide how much to trust the words you're reading.
What the AI does
After the math is finished, a language model receives the structured result and writes English (or Russian, or Spanish) prose around it. Its job is exclusively the prose layer:
- Find a good first sentence for your daily forecast.
- Phrase the meaning of "Saturn squares your natal Moon" without the doom of a hundred old astrology books.
- Vary tone across day-week-month so they don't read identically.
- Translate the same prose for the locale you've chosen.
That's it. The astrology itself — which planet is where, which aspect at what orb, which house cusp falls in which sign — is computed by Swiss Ephemeris before any AI sees anything.
What it never does
- It does not calculate the chart. The geometry is fixed before it gets involved.
- It does not invent aspects, transits, or planetary positions. If the prompt didn't contain it, the prose can't claim it.
- It does not predict specific future events. No "you will meet someone tall on Thursday".
- It does not give medical, mental-health, financial, or legal advice. The system prompt explicitly forbids it.
- It does not see your name, email, birth date, or location. Only anonymized structured astrological data is sent.
- We do not train the models on your data. The provider we use has training disabled on our account.
Models we use
We route through OpenRouter to specific models, picked for the job:
- Claude Haiku 4.5 — short daily and weekly forecasts. Fast, inexpensive, sharp on tone.
- Claude Sonnet 4.6 — full natal-chart interpretation and three-card tarot spreads. More space to reason, deeper synthesis.
We choose specific frozen versions. If we change models, we'll note it here.
What data it sees
The prompt sent to the model contains, roughly:
- Your computed astrological positions (planet, sign, degree, house).
- The aspects between them, with orbs.
- Current transits for the period being asked about.
- The locale of the requested output (en / ru / es).
- The reading type (daily, weekly, monthly, natal, synastry, tarot).
The prompt does not contain: your name, your account email, your birth city's full text, your IP address, or any other identifier. Two users with identical charts would get prose drawn from the same prompt.
When it gets it wrong
Language models occasionally write nonsense, or contradict themselves, or use a turn of phrase that lands badly. When you spot it, please tell us — there's a flag button next to every generated text in the app, and it goes straight to a real human inbox. We use those flags to refine our prompts.
We don't claim the prose is perfect. We claim it's honest about what it is.
If you'd prefer not to use AI-generated text at all, you can turn it off in Settings → AI features. The app then falls back to a curated library of human-written interpretations — narrower in scope, but entirely free of language models.