An astrology app has two halves. The first half is the app people touch: screens, sign-in, payments and the store listing. The second half is the astrology: the math that places the planets, the Hindu calendar, matching and the written readings. Build the first half. Buy the second half, unless the astrology math is the thing you sell.
Why split it that way? The app half is standard work for any app team. The astrology half is a specialty, and its mistakes are hard to spot. A wrong chart looks exactly like a right one. Our own engine took 265 public rules and 1,008 automated tests. It also took fixes for 3 cases that are easy to miss. They are clock changes, births with no sunrise, and Saturn stepping back across a sign line.
So split the work before you ask for a quote. Ask the company to quote the app half, and give them an astrology API for the rest. Here is each part, and 3 hard cases that show why.
The parts of an astrology app
Here are 13 parts an astrology app may need.
Part | What it does | Build or buy |
|---|---|---|
Chart engine | Finds each planet and the rising sign (lagna) at the birth moment | Buy |
Planet data (ephemeris) | Where each planet is on each date | Buy, with the engine |
Panchang and festivals | The Hindu calendar for a day and a city, with festival dates | Buy |
Kundli matching | The 36-point match between two birth charts | Buy |
Readings in plain words | Answers about personality, career, marriage and money | Buy |
Hindi | The same readings for Hindi speakers | Buy |
Printable report | A long reading users can save as a PDF | Buy |
Place search | Turns "New York" into coordinates and a time zone name | Build, or use a map service |
Screens and design | Everything users see and tap | Build |
Accounts | Sign-in, profiles and saved birth details | Build |
Payments | Subscriptions, purchases and refunds | Build |
Notifications | Daily reminders and alerts | Build |
Store listing | Your App Store and Google Play pages | Build |
The split is about where your app is different. Your screens, brand and prices make your app yours. The planet positions for Aug 15, 1990, are the same in every app, or they should be.
Build the app half: it is normal app work
Screens, accounts, payments and alerts are standard work. Every app with paying users needs them, astrology or not. Put your budget here, because users see and judge it. The API returns data, and your app decides how to show it.
Call the API from your server. A key placed inside a mobile app can be pulled out of it. Keep the key on your server, and let the app talk to your server. The authentication guide lists the other key habits.
Place search is the one gray area. The API needs latitude, longitude and a time zone name, such as America/Chicago for Jane Doe in Chicago. Ask for the zone name, not a fixed offset like UTC-6 (6 hours behind world time). The first hard case shows why.
Buy the astrology half: what our own build took
We built the engine behind this API. We measured these numbers from our own code on Oct 11, 2026.
265 public rules, each with an ID. Every rule the API follows is in our rulebook. In 143 of them, traditions differ, so we had to pick one and say which.
1,008 automated tests in 56 files. Several exist because of real bugs we found and fixed.
59 calculation files with 14,759 lines of Python. That is before a single word of reading text.
Checked against Swiss Ephemeris, an outside reference. We tested 11 charts from 1855 to 2026. On each, the Sun, the Moon and the five planets are within 1.5 arcseconds, or 1/2,400 of a degree. See how accurate the calculations are.
The edge cases show where the effort went.
Hard case 1: the hour that happens twice
A forgotten clock change moves the whole chart. John Doe was born on Aug 15, 1990, at 10:30 AM in New York, during daylight saving time. With the zone name America/New_York, our API puts his rising sign at Virgo 19°56'. With a fixed UTC-5, it becomes Libra 1°48'. Then every planet lands in a different house (one of the 12 areas of life).
Some times happen twice. On Oct 28, 1990, New York set its clocks back at 2:00 AM, so 1:30 AM happened twice. A baby born then could have a rising sign of Leo 0°36' or Leo 12°24'. Our API does not guess. It asks which one, through a field called timezone_fold.
Some times never happen. On Apr 1, 1990, the clocks jumped from 2:00 AM to 3:00 AM. A birth at 2:30 AM that day is a mistake, and the API says so.
Your app team builds one extra question, shown only when the API asks: "Was this before or after the clocks went back?"
Hard case 2: a birth with no sunrise
A Hindu day starts at sunrise. Much of the panchang (the Hindu calendar for a day) counts from that moment. But in Tromsø, Norway, the Sun never sets on June 21, so there is no sunrise.
Our engine says so instead of making one up. Ask for the daily panchang in Tromsø on June 21, 1990, and the API returns a "no sunrise" error code. The birth chart for that day still works. The birth details reading still answers, minus the parts that need a sunrise.
This one was a real bug. Before our fix, a birth like this made the full report and the birth details reading fail with a server error.
Hard case 3: Saturn steps back across a sign line
Sade sati is a demanding stretch of about 7.5 years. Your Moon sign is the sign the Moon was in at birth. Sade sati runs while Saturn passes the sign before it, the sign itself and the sign after. Saturn sometimes seems to move backward for a few months, which astrologers call retrograde. So it can cross one sign line 3 times.
Here are John Doe's real dates. His Moon is in Taurus, so his next sade sati starts when Saturn enters Aries on Jun 2, 2027. Saturn steps back out on Oct 19, 2027, and returns on Feb 23, 2028. It reaches Taurus on Aug 8, 2029, steps back on Oct 5, 2029, and returns on Apr 16, 2030. The stretch ends on Jul 12, 2034.
A tool that looks for one entry and one exit gets this wrong. It shows 2 separate sade satis, or one that starts too late. Our API returns one stretch, with the break and every phase change dated. Our tests check Saturn's real 2022 path the same way. It entered Aquarius in April, stepped back out in July and came back for good in January 2023.
Try the chart math before you write the brief. Enter John Doe's birth details in our free kundli tool. API accounts open on November 26, 2026. Get an email when accounts open
What to put in your brief to the development company
A short brief for the astrology part saves a round of questions.
Name the API call for each screen. Chart screen:
POST /v1/planets. Daily screen:POST /v1/reading/daily. Match screen:POST /v1/matching. Paid report:POST /v1/report.Ask for server-side calls and a place search that returns a zone name.
Ask how they handle the 2 clock replies. One is the time that happens twice. The other is the time that never happened.
Ask them to show the inputs back, including the UTC offset, so users catch mistakes.
Give them a test birth. John Doe (Aug 15, 1990, 10:30 AM, New York) should show Virgo rising at 19°56'. His Moon should be in Taurus at 24°35'.
The docs list every call, and the astrology apps use case maps them to screens. For an AI chat, see the AI assistants use case and how to build an AI astrologer that gets the chart right.
What the API covers today
The API has 50 endpoints (kinds of request) under /v1/. They cover birth charts and 16 divisional charts (extra charts cut from the birth chart). Dashas (planetary periods), the panchang, festivals, matching and transits are covered too. You also get 27 topic readings, a daily reading and a printable report, in English, Hindi and more languages as they launch. Dates run from 1850 to 2149. Plans are on the pricing page.
When building your own engine makes sense
Build the astrology half if the math is your product. Maybe you have a new method, or you need a tradition we do not cover. Our API is Vedic only. Our rulebook also names 14 gaps: rules people expect that the API does not compute yet.
Then plan for the work we did. You will write a rule wherever traditions disagree, and test against an outside reference. You will also meet the same hard cases we did.
Brief the app half now, and connect the astrology at launch. API accounts open on November 26, 2026. Join the launch list
For everyone else, the split is simple. Users judge your app by its screens and by whether their chart is right. Spend your build budget on the screens. Buy the chart from a team that has already met the hard cases.
