Market holidayJapan will be closed on 2026-11-03 (Culture Day), 5 days from now.All announcements

API latency self-test

From your browser, send sequential requests to api.tiqex.io and measure round-trip time (RTT) with P50 / P95 / P99 and a distribution chart. TIQEX endpoints only—no cross-vendor comparison and no API key.

Test parameters

All three endpoints are public and return 200—no API key needed; this measures browser-to-edge round trip, excluding market data lookup.

3–30 sequential requests

Request URL:https://api.tiqex.io/api/v1/plans

Results

Not started
—P50 median
—P95
—P99
—Mean
—Jitter (std dev)
—Success / total

Latency distribution

Per-sample

Per-sample details
#Time (ms)Result
No data
With few samples P99 ≈ max; for stable tail latency, take 30+ samples and re-test from your servers.

How to use

  1. What is measuredTotal time from browser request to response: network RTT + TLS + gateway. Excludes matching and is not WebSocket push latency.
  2. Why drop the 1st sampleThe first request pays DNS and TLS handshake, often much slower; production long-lived connections do not pay that every time.
  3. Reading percentilesP50 for typical feel, P95 / P99 for tail jitter; set alert thresholds from P95 / P99, not the mean.
  4. If requests failExtensions, corporate proxy, or firewall may block requests. Retry on another network or use curl -w "%" on a server.

Related API

Live data over WebSocket, not polling

A free API Key can call realtime crypto quotes.