> For the complete documentation index, see [llms.txt](https://docs.premium-positioning.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.premium-positioning.com/learn/rtktest.md).

# RTKTEST

RTKTEST.com, the open benchmark Premium Positioning runs: how a session works, the composite score, the public leaderboard, and the methodology partners Geo++ and the Leica GNSS Spider team.

RTKTEST.com, the open benchmark Premium Positioning runs, measures correction services against a surveyed point. It streams up to three services to a fixed reference receiver, called a probe, and compares every position with the known coordinate. RTKTEST.com has been open to the public since 24 August 2026.

Many providers claim 2 to 3 cm, and RTKTEST is where that claim gets checked. RTK Premium is tested under the same methodology as every other service, with its raw data open to the same verification.

### How a session works

1. Choose the probe closest to where you operate. A probe is a fixed reference receiver at coordinates surveyed to sub-centimetre accuracy.
2. Enter the NTRIP details for up to three services: caster address, port, mountpoint, username and password.
3. Run the session. For 15 minutes the probe processes every stream from the same antenna and the same satellite observations, and computes one position per second for each service.
4. Read the results. They are sent to the email address you entered and stay private to it.

Because every service sees the same sky, atmosphere and multipath at the same moment, a difference in the result comes from the correction service. Rate limits, engine settings and scoring are identical for every service.

Your credentials sign in directly with each provider's caster and are not stored. RTKTEST identifies a service by its caster address and mountpoint.

### Composite score

A session reports trueness, precision, fix availability and time to first fix, the measures defined in [Testing a correction service](https://docs.premium-positioning.com/learn/testing-a-correction-service.md#what-to-record). The composite score weights them as follows:

| Measure           | Weight |
| ----------------- | ------ |
| Accuracy          | 40%    |
| Fix availability  | 25%    |
| Precision         | 25%    |
| Time to first fix | 10%    |

Correction age is reported as a health flag above 5 seconds and is not scored. Every session returns the scores at fixed epochs, the full distributions, a PDF report and the raw RINEX, so you can replay the session and check the numbers yourself. The [RTKTEST methodology](https://www.rtktest.com/methodology) publishes the scoring curves, the open-source engine and its configuration.

### Probes

Public probes take streams from open networks such as EUREF and IGS and use a published coordinate, so anyone can replay the procedure. Partner probes are private operator stations. The probe map shows the two types in different colours, and each result states which type it was measured on. In North America, RTKTEST measures in NAD83(2011), epoch 2010.0, the frame RTK Premium delivers in the United States. See [Reference frames and constellations](https://docs.premium-positioning.com/learn/reference-frames-and-constellations.md).

### Methodology partners

Geo++ and the Leica GNSS Spider team at Leica Geosystems both build software for GNSS reference networks, and both have put their name behind how the test is run. A partnership gives no measurement advantage: networks operated or equipped by a partner are scored on the same terms and stay anonymised on the leaderboard.

#### Geo++

We partner with Geo++, the industry reference for GNSS quality, to further develop our evaluation methodology and enable even more rigorous, data-driven analysis of GNSS correction services.

> "Transparent measurement helps users understand what they are buying, and it helps providers understand where their networks can improve. That benefits everyone in the industry."
>
> Temmo Wübbena, Managing Director, Geo++

#### Leica GNSS Spider team

The Leica GNSS Spider team at Leica Geosystems is a partner on the evaluation methodology behind RTKTEST.

> "Coverage maps and fix rates tell you very little about actual positioning quality within a GNSS network. RTKTEST measures what matters: simultaneous observations, identical position calculation, and raw data open to independent verification. That is the standard the industry should hold itself to, and why we put our name behind the methodology."
>
> Markus Roland, Senior Product Manager, GNSS Networks & Reference Stations, Leica Geosystems AG

### Testing RTK Premium

Enter the credentials from your trial or subscription with the caster address rtkpremium.xyz, port 2101 and mountpoint COMMON. No account yet? Start a [14-day free trial](https://www.premium-positioning.com/start-14-day-free-trial?utm_source=docs\&utm_medium=referral\&utm_campaign=docs\&utm_content=rtktest) to get your credentials.

Every RTKTEST session also measures RTK Premium as a fixed reference service. Its figures appear in a separate block of the result, are never ranked against the services you chose, and are kept out of the public leaderboard. When you enter RTK Premium as one of your own services, it is ranked like any other.

### Public leaderboard

The public leaderboard shows networks as anonymised numbers, reshuffled with every publication. A network appears for a region only after 40 independent sessions there within the quarter, and sessions on beta probes do not count. An operator can claim its listing after verifying ownership and consenting to publication. Individual sessions are never published or attributed.

### Limitations

* A probe measures a service at its own location, not across the whole service area.
* Ionospheric conditions vary, so a single session gives an indicative rank. Repeat sessions on different days before you decide.
* A VRS stream is built for the probe position, which can sit several kilometres from where you work. To measure at your own sites, follow [Testing a correction service](https://docs.premium-positioning.com/learn/testing-a-correction-service.md).

### FAQ

Who runs RTKTEST? Premium Positioning runs it. The methodology is public and versioned, every service is measured with the same engine and settings, and the raw data is archived so a third party can recompute any result.

Who are the RTKTEST methodology partners? Geo++ and the Leica GNSS Spider team at Leica Geosystems are partners on the evaluation methodology behind RTKTEST.

Does RTKTEST keep my NTRIP password? RTKTEST does not keep it. Your credentials sign in directly with your provider's caster and are not stored.

Can I compare my current provider with RTK Premium? You can. Enter both in the same session, and they are measured on the same probe over the same 15 minutes.

Is there a probe near me? The probe map opens when you start a session. Coverage depends on where suitable reference stations are available, so pick the probe closest to where you operate. To ask for a probe in your region, email <info@premium-positioning.com>.

Last updated: 15 October 2026
