MiyaIP 简体中文 free tools

MiyaIP proxy link measurement

Proxy Speed Test
Measure the link behind every millisecond

Use 7 sets of formal measurements to observe server-to-proxy TCP connection time, first byte time, multi-connection download rate, and failure rate. The results are a measurement of the current link, not a long-term stability commitment.

My public IP Recognizing… Waiting

It is only used to compare proxy exports and does not participate in speed rating; failure to identify will not affect the speed measurement.

Test one proxy at a time. Supports hostname, public IPv4, square bracket IPv6 and authentication credentials.

support host:port, Host:Port:Username:Password and socks5://user:pass@host:port.

Automatic mode performs lightweight protocol probing first; the protocol prefix in the address takes precedence.

Only metrics are saved, not proxies and credentials

Last 8 Speed Tests

After completing a Speed Test, the TTFB and download rate trends will be displayed here.

Fixed process, repeatable comparison

How the Speed Test Works

All targets are preset by the server, and the proxy cannot direct tools to intranet addresses. Lightweight delay samples are collected sequentially, and multi-connection aggregation is used in the download phase to reduce underestimation caused by small files and TCP slow start.

  1. 1

    Verify proxy host

    Resolve DNS, reject any non-public results, and pin subsequent connections to the verified address.

  2. 2

    Confirm agency agreement

    Automatic mode tries common protocols with lightweight requests, and authentication errors will not continue to be retried meaninglessly.

  3. 3

    Collect 6 TTFB samples

    Two fixed HTTPS probes at different network locations are executed for 3 rounds each, and the connection and first byte time are recorded.

  4. 4

    Perform multiple connection downloads

    By default, 4 download streams are established at the same time, each with 2 MiB; the aggregate rate is calculated based on the actual total bytes of successful streams and the common transmission time.

The conclusion can be explained

How Speed Ratings Are Assigned

Binning takes into account completeness, TTFB and download rate simultaneously. The connection time is greatly affected by the deployment location and is only shown as a fact.

Fast

All 7 groups succeeded

Average TTFB is no more than 300 ms and download rate is at least 500,000 B/s.

Average

At least 5 groups successful

Average TTFB does not exceed 800 ms and download rate is at least 100,000 B/s.

slow

Key indicators are complete

Able to complete valid measurements, but not all conditions for "fast" or "average" are met.

Not binning

Insufficient sample

With less than 3 TTFB samples or downloaded samples not available, don’t force conclusions with unknown data.

FAQ

Correct understanding of speed results

Is this the speed from my computer to the proxy?

No. Requests are made by the MiyaIP application server, and the results reflect the links between the application server, proxy, and fixed probes. Your local carrier, Wi‑Fi and device performance are not part of the measurement path.

Why not use the name Ping?

HTTP and SOCKS proxies generally cannot forward ICMP. The tool measures the connection time to the proxy TCP port and measures the time to first byte separately to avoid conflating different metrics.

Why are the results different every time for the same proxy?

Inter-network routing, proxy load, target congestion, and transient packet loss all cause fluctuations. It is recommended to test several times at similar times and judge based on the minimum value, average value, maximum value and failure rate.

Why are the results different from the Speedtest results in the browser?

Regular Speedtest measures your device to a nearby Speed Test server, usually adaptively increasing the amount of data and using multiple connections. This tool measures the link from the current MiyaIP application server to the fixed download probe via proxy; the test starting point, target, routing and proxy exit may be different, so the values ​​cannot be directly equivalent.

Will the upload speed be measured here?

The first version only tests the server download through the proxy, and will not upload test data. The "50 Mbps upstream and downstream" you see in other Speedtests includes upload capabilities, but this page only shows this server-side download link.

How much proxy traffic will Speed Testing consume?

The default download phase is 4 channels × 2 MiB, and all successes consume approximately 8 MiB; the other 6 requests only read lightweight responses, and automatic protocol recognition may generate a small amount of additional traffic.

Will the proxy account number and password be saved?

No. Credentials are only used for this server connection. Responses, exports, and recent history do not contain credentials; recent history does not save the proxy host.