››
›
Every plan has two limits: requests per month and requests per minute. This guide explains how they interact, what counts toward your quota, how to estimate usage, and how to cache aggressively to multiply your effective capacity.
Every Satta Matka API plan has two independent limits: requests per month (your monthly quota) and requests per minute (your concurrency cap). Hitting either one returns a 429 Too Many Requests response. This guide explains how the limits work, how to estimate your usage before you upgrade, and how to multiply your effective capacity 5–10x with caching.
| Plan | Requests / month | Requests / minute | |---|---|---| | Free Trial (2 days) | 1,000 | 10 | | Basic (₹99/mo) | 30,000 | 60 | | Pro (₹479 / 6 mo) | 200,000 | 200 | | Enterprise | Custom | Custom |
The monthly quota is a hard counter — every successful API request (200 response) decrements it by 1. The per-minute cap is a sliding window — if you exceed it, the API returns 429 and the request does NOT count against your monthly quota.
So you are billed for "your fault" errors (bad request, wrong auth, wrong slug) but not for "our fault" errors. Be careful with 404s — a typo in a market slug on a busy page can burn through your quota fast.
A typical matka results site makes:
/api/results/live (the board).If you have 1000 visitors/day, each staying 3 minutes on average and polling every 30s:
1000 visitors × (1 + 3 min × 2 polls/min) = 1000 × 7 = 7,000 requests/day
That is 210,000 requests/month — Pro plan territory. Without caching.
You can cut that number by 80%+ with two layers of caching:
The live board response is identical for every visitor for ~30 seconds. Set a Cache-Control: max-age=30, s-maxage=30 header on your server-side proxy and the user's browser will reuse the same response for 30 seconds, skipping the API call entirely.
s-maxage)If your site is behind a CDN (Cloudflare, Vercel Edge, Fastly), set s-maxage=30 (or higher) on the response. The CDN will serve the cached response to thousands of users from a single API call. This is the single biggest lever for matka result sites — a 30-second CDN cache turns 33 req/s into 0.03 req/s at your origin.
If you proxy the API through your own server (recommended, so you don't expose your key), add a tiny in-memory cache that serves the last-fetched response for 10 seconds. This dedupes concurrent requests — if 50 users hit your origin in the same second, only one of them triggers an API call; the other 49 get the cached response.
let cache = null
let cacheAt = 0
async function getLiveBoard() {
if (cache && Date.now() - cacheAt < 10000) return cache
cache = await fetch('https://sattamatkaapi.live/api/results/live', {
headers: { Authorization: 'Bearer ' + API_KEY }
}).then(r => r.json())
cacheAt = Date.now()
return cache
}
If your page is in a background tab, you do not need to poll. Use the visibilitychange event:
let pollInterval
document.addEventListener('visibilitychange', () => {
if (document.hidden) {
clearInterval(pollInterval)
} else {
refresh() // immediate refresh on tab focus
pollInterval = setInterval(refresh, 30000)
}
})
This typically cuts polling traffic by 30–50% on sites where users open a results tab and leave it in the background.
If you have a backend, switch from polling to webhooks (see Setting Up Webhooks for Auto Result Updates). Your server gets one POST per declared result — roughly 550 events per day across all 276 markets. That is 16,500 requests/month, regardless of how many users you have. Compare that to 210,000 requests/month for 1000 polling visitors, and you can see why webhooks are dramatically cheaper at scale.
If you need to backfill a market's history (e.g. on first integration), do not call /api/results/{slug} 30 times for the last 30 days. Use the bulk history endpoint:
curl -H "Authorization: Bearer $MATKA_API_KEY" \
"https://sattamatkaapi.live/api/results/history/kalyan?days=30"
One request returns 30 days of data. That is 30x cheaper than per-day calls.
When you do hit the rate limit, your client should:
Retry-After header (seconds to wait before retrying).async function safeFetch(url, opts) {
const res = await fetch(url, opts)
if (res.status === 429) {
const wait = parseInt(res.headers.get('Retry-After') || '5', 10)
await new Promise(r => setTimeout(r, wait * 1000))
return safeFetch(url, opts)
}
return res
}
Retry-After.A typical Pro plan (200k requests/month) can comfortably serve 5000+ daily visitors with proper caching. Without caching, you would need Enterprise.
Get 2 days of full access to every endpoint — live results, history, charts, webhooks. No credit card required.
View pricing & start trial