Vedikks calculates a sidereal chart with the Swiss Ephemeris library using the Lahiri ayanamsa, hands that chart plus your relevant earlier messages to a language model, and writes the answer in plain English. The astronomy is checkable. The interpretation is not, and this page says so.
- Exact library, flags and ayanamsa published
- Three reference charts you can recompute
- Every layer has a stated limit
- Corrections go to a named person
Last checked against the running system on 25 August 2026. Maintained by Mrigang Singh, founder. Corrections: support@vedikks.com.
1. The calculation layer
Every chart is computed server-side. Nothing about the astronomy is generated by a language model, and if the ephemeris module fails to load, the chart endpoint returns an error rather than a guess.
| Component | What Vedikks uses |
|---|---|
| Ephemeris library | swisseph (Node binding to the Swiss Ephemeris), version 0.5.17 |
| Ephemeris mode | SEFLG_MOSEPH — the Moshier analytical ephemeris built into the library, not external JPL or .se1 data files |
| Zodiac | Sidereal (SEFLG_SIDEREAL) |
| Ayanamsa | Lahiri / Chitrapaksha, set once at start-up with swe_set_sid_mode(1, 0, 0) |
| Planets | Sun, Moon, Mercury, Venus, Mars, Jupiter, Saturn, computed with swe_calc_ut |
| Lunar nodes | Mean node (SE_MEAN_NODE) for Rahu; Ketu is taken as Rahu + 180° |
| Ascendant | swe_houses(jd, lat, lng, 'P') gives a tropical Placidus ascendant; the Lahiri ayanamsa for that Julian Day is then subtracted to get the sidereal lagna |
| Time handling | utHour = hour + minute / 60 − utcOffset, then swe_julday(…, SE_GREG_CAL) |
| Cross-reference | Tropical Sun, Moon and ascendant are also computed, so a Western-zodiac comparison is possible |
Where this differs from other apps — and why your numbers may not match
Mean node, not true node. Apps that use the true node place Rahu and Ketu up to roughly 1.5° away from ours. That is enough to change a nakshatra pada, and occasionally a nakshatra, near a boundary.
Lahiri, not Raman or KP. Ayanamsa systems disagree by about a degree. Two apps using different ayanamsas will put a planet born near a sign cusp in different signs. Neither is a bug; they are different conventions. We publish ours so the disagreement is legible.
Moshier mode. The Moshier ephemeris agrees with JPL to well under an arcsecond for modern dates. It is immaterial for signs, nakshatras and dashas, and only visible if you are comparing raw longitudes to many decimal places.
The UTC offset is taken from your entry, not from a historical timezone database. For births in India from 1955 onwards, IST is +05:30 and this is safe. For earlier Indian births, or births in countries that observed daylight saving, the offset should be checked. A wrong offset moves the lagna, and a wrong lagna moves every house. If your chart looks wrong, this is the first thing to check.
Reference charts you can recompute
These three charts use invented birth data, not any real person's. They were produced by the same library, flags and ayanamsa listed above, on 25 August 2026. If you run the code below and get these numbers, our calculation layer is doing what this page says. If you run another app and get different numbers, the table above explains the likely reason.
Machine-readable copy: reference-charts.json.
REF-A — 1 January 2000, 12:00:00, UTC+05:30, New Delhi (28.6139°N, 77.2090°E)
JD(UT) = 2451544.770833 Lahiri ayanamsa = 23.857084°
| Body | Sidereal longitude | Sign and degree | Nakshatra |
|---|---|---|---|
| Ascendant | 343.1853° | Pisces 13°11'07" | Uttara Bhadrapada pada 3 |
| Sun | 256.2821° | Sagittarius 16°16'56" | Purva Ashadha pada 1 |
| Moon | 196.7117° | Libra 16°42'42" | Swati pada 4 |
| Mars | 303.9323° | Aquarius 3°55'56" | Dhanishta pada 4 |
| Mercury | 247.6796° | Sagittarius 7°40'46" | Mula pada 3 |
| Jupiter | 1.3906° | Aries 1°23'26" | Ashwini pada 1 |
| Venus | 217.4355° | Scorpio 7°26'08" | Anuradha pada 2 |
| Saturn | 16.5470° | Aries 16°32'49" | Bharani pada 1 |
| Rahu (mean) | 101.1996° | Cancer 11°11'58" | Pushya pada 3 |
| Ketu (mean) | 281.1996° | Capricorn 11°11'58" | Shravana pada 1 |
REF-B — 15 June 1990, 05:45:00, UTC+05:30, Mumbai (19.0760°N, 72.8777°E)
JD(UT) = 2448057.510417 Lahiri ayanamsa = 23.723717°
| Body | Sidereal longitude | Sign and degree | Nakshatra |
|---|---|---|---|
| Ascendant | 55.3279° | Taurus 25°19'40" | Mrigashira pada 1 |
| Sun | 59.9347° | Taurus 29°56'05" | Mrigashira pada 2 |
| Moon | 315.1309° | Aquarius 15°07'51" | Shatabhisha pada 3 |
| Mars | 346.9619° | Pisces 16°57'43" | Revati pada 1 |
| Mercury | 41.1288° | Taurus 11°07'44" | Rohini pada 1 |
| Jupiter | 82.0562° | Gemini 22°03'22" | Punarvasu pada 1 |
| Venus | 24.4747° | Aries 24°28'29" | Bharani pada 4 |
| Saturn | 270.3333° | Capricorn 0°20'00" | Uttara Ashadha pada 2 |
| Rahu (mean) | 285.9965° | Capricorn 15°59'48" | Shravana pada 2 |
| Ketu (mean) | 105.9965° | Cancer 15°59'48" | Pushya pada 4 |
Note on REF-B. The Sun sits at Taurus 29°56'05", four minutes of arc from Gemini. This case is deliberately included: it is exactly where an app on a different ayanamsa, or a birth time off by a few minutes, will report a different sun sign. If you want one test that exposes convention differences between astrology apps, use this one.
REF-C — 9 November 1975, 23:10:00, UTC+05:30, Chennai (13.0827°N, 80.2707°E)
JD(UT) = 2442726.236111 Lahiri ayanamsa = 23.519839°
| Body | Sidereal longitude | Sign and degree | Nakshatra |
|---|---|---|---|
| Ascendant | 102.2028° | Cancer 12°12'10" | Pushya pada 3 |
| Sun | 203.1768° | Libra 23°10'36" | Vishakha pada 1 |
| Moon | 281.5344° | Capricorn 11°32'04" | Shravana pada 1 |
| Mars | 69.0663° | Gemini 9°03'59" | Ardra pada 1 |
| Mercury | 191.8822° | Libra 11°52'56" | Swati pada 2 |
| Jupiter | 352.8120° | Pisces 22°48'43" | Revati pada 2 |
| Venus | 156.6386° | Virgo 6°38'19" | Uttara Phalguni pada 3 |
| Saturn | 99.4285° | Cancer 9°25'43" | Pushya pada 2 |
| Rahu (mean) | 208.5116° | Libra 28°30'42" | Vishakha pada 3 |
| Ketu (mean) | 28.5116° | Aries 28°30'42" | Krittika pada 1 |
Reproduce it
Node 18 or later, then npm install swisseph@0.5.17:
const swe = require('swisseph');
swe.swe_set_sid_mode(1, 0, 0); // SE_SIDM_LAHIRI
// REF-A: 2000-01-01 12:00 IST, New Delhi
const utHour = 12 + 0 / 60 - 5.5;
const jd = swe.swe_julday(2000, 1, 1, utHour, swe.SE_GREG_CAL);
const ay = swe.swe_get_ayanamsa_ut(jd);
const SID = swe.SEFLG_MOSEPH | swe.SEFLG_SIDEREAL;
console.log(swe.swe_calc_ut(jd, swe.SE_SUN, SID).longitude); // 256.2821
console.log(swe.swe_calc_ut(jd, swe.SE_MOON, SID).longitude); // 196.7117
const asc = swe.swe_houses(jd, 28.6139, 77.2090, 'P').ascendant;
console.log(((asc - ay) % 360 + 360) % 360); // 343.1853
Longitudes above are rounded to four decimal places. If your build of the library reports a value that differs in the fourth decimal, that is rounding, not disagreement. If it differs in the first decimal, something in the inputs is different — usually the ayanamsa setting or the UTC offset.
2. The interpretation layer
Once the chart exists, a language model (Claude, from Anthropic) writes the answer. The chart is computed first and passed in; the model is not asked to do astronomy, and it does not calculate positions.
What the model receives before it replies: your calculated placements and lagna, your Vimshottari dasha position, current transits relevant to the question, the astrologer persona you selected if you chose one, and the recent messages of the conversation you are in.
What “accurate” can and cannot mean here
It can mean: the planetary positions match the published reference charts; the dasha arithmetic is right; the answer cites placements that are actually in your chart rather than invented ones.
It cannot mean: that a prediction about your life will come true. Swiss Ephemeris is an astronomy library. It can tell you where Saturn was to the arcsecond. It cannot validate a claim about your career in 2027, and no amount of calculation precision changes that.
Where traditions disagree, we do not hide it. Jyotish has genuine internal disputes — which reference point to read career from, whether a divisional chart or the navamsa of a lord takes priority, whether a dosha applies. Where the classical sources conflict, the honest answer says so rather than picking one and sounding certain. Our blog writes several of these disputes up in detail.
3. Your data
What is stored, stated plainly rather than summarised as “secure”.
- Your birth details, per profile, so the chart can be recalculated rather than cached blindly.
- The questions you ask and the answers you receive, saved with the profile they were asked about.
What you cannot do from inside the app yet
- Browse or export everything stored about you.
- Remove your first (“Self”) profile. Additional profiles you create can be removed, with a typed confirmation.
Full erasure of your account and its stored records is handled by our team on request to support@vedikks.com, within the window set out in the privacy policy. The processors involved — where the database sits, who generates the text, what analytics run — are named in that policy.
If you only want a chart and would rather store nothing, the free Kundli generator calculates it in your browser without an account.
4. Safety and limits
Vedikks is guidance and reflection. It is not a prediction to plan a life around, and it is not care.
- No guaranteed outcomes, no fixed dates for marriage, jobs, recovery or money.
- Not a substitute for medical, mental-health, legal or financial advice, and not a diagnostic tool. If a question is really a health question or a money question, it needs a professional, not a chart.
- Not a remedy shop. Vedikks does not sell gemstones, pujas or fear-priced fixes, and it should not manufacture dread about Saturn, Rahu, Mangal dosha or Sade Sati in order to keep you reading.
- Not a reassurance machine. Asking the same worry ten different ways is a pattern worth noticing rather than feeding; we have written about that here.
If you are in distress, please talk to someone who can actually help — a doctor, a therapist, or someone you trust. An astrology app is not the right place to carry that alone.
5. Change log and corrections
This page is versioned deliberately. When the calculation stack, data handling or the stated limits change, the change is listed here with a date rather than quietly edited in.
- 2026-08-25 First publication. Calculation stack, reference charts REF-A to REF-C, data handling, and safety limits documented and checked against the running system.
Found something wrong?
If a number on this page does not reproduce, or the product does not behave the way this page describes, that is a defect in the page or the product and we want to know. Write to support@vedikks.com. Corrections are owned by Mrigang Singh, founder of Vedikks. Substantive corrections get added to the change log above with the date they were made.