Kundli matching API for matrimony platforms
36-guna ashtakoot, Mangal dosha with cancellation rules, and batch matching — as clean JSON.
What a matrimony product needs
- 36-guna matching —
POST /v1/milan/ashtakootreturns all 8 koots with points, total/36, percentage and verdict (11 Milan endpoints) - Mangal dosha honesty — detection with classical cancellation rules, so profiles aren't wrongly labelled manglik
- Batch screening — Batch Milan scores up to 1,000 person1/person2 pairs per synchronous call, more via webhook (repeat one profile as person1 to screen it against many), built for match-suggestion pipelines
- Missing birth times — tob is optional (a sunrise chart is used); Milan responses don't carry an accuracy flag, so badge matches computed without a birth time as approximate in your UI
- Bilingual where it matters — every text field has an English value and a Hindi
_hitwin (e.g."verdict": "Average","verdict_hi": "सामान्य"), and Moon signs carry amoon_sign_hitwin; koot labels (varna, yoni, gana, nadi) are English
Integration sketch
POST /v1/milan/ashtakoot
{"person1": {"dob": "1990-06-15", "tob": "10:30", "lat": 28.6139, "lon": 77.209, "tz": 5.5},
"person2": {"dob": "1992-11-22", "tob": "08:20", "lat": 26.85, "lon": 80.95, "tz": 5.5}}
→ {"success": true,
"data": {"total": 20.0, "max": 36, "percentage": 55.6, "verdict": "Average", "verdict_hi": "सामान्य", "nadi": {"score": 8, "max": 8, "nadi_dosha": false, ...}, ...},
"meta": {"ayanamsa": "Lahiri", ...}}The values shown are what a sk-test- key returns — test mode always computes its fixed sample pair, not your input. Develop the whole flow on free unlimited sk-test- keys; the structure is guaranteed identical in production. Pricing is flat per call — never a credit multiplier, and no overage bills. Once a month's quota is used up, calls return a 429 until the 1st, so watch the X-Quota-Remaining header and move up a plan ahead of festival season if you need the headroom.
Can it handle profiles without birth time?
Yes — tob is optional; the chart is computed from sunrise. Milan responses don't carry an accuracy flag, so badge such matches as approximate in your UI (the Moon can change nakshatra within a day, which changes several koots).
How does batch matching price?
Batch Milan endpoints count one call per request like everything else — there is no credit multiplier. See the pricing page for per-plan volumes.
Is Mangal dosha detection fear-free?
It reports the dosha AND the classical cancellation (parihara) rules that apply, so a technically-manglik profile with cancellations is shown honestly — important for matrimony UX and fairness.
Time to bring the cosmos into your app.
Start free. No credit card. API key in under 1 minute.