A new and highly experimental digital mode called FT2 has just entered its first on-air testing phase. On February 16, 2026 at 22:47 UTC, a small group of Italian radio amateurs reported the first successful signal exchanges using a modified WSJT-X codebase aimed at one goal: extreme-speed QSOs.
This is not an official WSJT-X release and not a finalized protocol. It is an early “in-the-wild” experiment based on a fork of WSJT-X v3.0.0-rc1, named Decodium 3 v3.0.0-rc1 intended to evaluate how far transmission timing can be compressed when signals are strong and conditions are favorable.
What’s different: QSOs in 7–11 seconds
During the first tests, stations IZ8VYF, IZ8XXE, and IC8TEM reportedly completed full contacts in roughly 7 to 11 seconds.
For comparison (typical end-to-end timing):
- FT8: ~60 seconds (15 sec /cycle) ~ 60 QSOs/h
- FT4: ~30 seconds (7.5 sec /cycle) ~ 120 QSOs/h
- FT2 (experimental): ~11 seconds (3.8 sec /cycle) ~ 240 QSOs/h
Technical snapshot (early test parameters)
FT2’s speed comes from a much shorter T/R window:
- Cycle length: ~3.75–3.8 seconds
- Modulation: 8-GFSK (as used by FT8/FT4)
- Payload: 77-bit (same message payload size class)
- Sensitivity (initial observations): decoding around -12 to -14 dB, implying higher SNR requirements than FT8
If these characteristics hold, FT2 could be particularly interesting for high-rate contesting scenarios and intense pile-ups, where speed matters more than weak-signal sensitivity.
FT2 operating frequencies (current test calling points)
Below are the frequencies currently used by testers for FT2 activity:
- 160m: 1.843 MHz
- 80m: 3.578 MHz
- 60m: 5.360 MHz
- 40m: 7.052 MHz
- 30m: 10.144 MHz
- 20m: 14.084 MHz
- 17m: 18.108 MHz
- 15m: 21.144 MHz
- 12m: 24.923 MHz
- 10m: 28.184 MHz
PSKReporter showing FT2 spots
Several FT2 contacts are already tracked by PSKReporter, this is the link to FT2 QSOs
Current status: limited alpha distribution
FT2 is available running DECODIUM 3, and it remains in a very early experimental stage. According to the testers, binaries are not known to be publicly available on standard repositories yet and are currently shared in a restricted circle (including via a private WhatsApp group) to keep feedback controlled while the protocol is still being refined. We will keep monitoring.
Project notes and next steps
The project is described as “Made in Italy,” developed by IU8LMC with support from the ARI Caserta Team. Early activity has reportedly been logged on 40m and 80m, with further refinements expected before any wider release.
More details—public availability, documentation, and a clearer technical specification—are expected in the coming weeks.
How to download FT2 Software
Currently the WSJT-X fork, can be downloaded via the DECODIUM page
FT2 Digital Mode Insights
FT2 Frequently Asked Questions
DECODIUM is basically a Fork of the WSJT-X open source software developed by K1JT. As happened for other forks develepments some improvements have been included in Release Candidates and then on Official Releases. Based on current development time of the WSJT-X project, we believe it will require some months.
Based on the experimental data and technical documentation, FT2 is designed as a specialized tool for scenarios where signal strength is high and time is the limiting factor. It is not intended to replace FT8 for weak-signal work but rather to maximize throughput when conditions allow.
Here are the ideal use cases for the 11-second QSO speed offered by FT2:
High-Rate Digital Contesting, DXpeditions, Special Event Stations, and Pile Ups in general.
FT2 is positioned as a potential “RTTY-replacement” for modern digital contesting. Because an FT2 contact takes only 11 seconds compared to FT8’s 60 seconds or FT4’s 30 seconds, it offers a significant strategic advantage in competitions.
• Throughput: It can triple the contact rate compared to FT8.
• Volume: Ideally suited for contests where the final score is determined by the sheer volume of contacts made in a limited time.
Based on the technical specifications provided in the experimental documentation, FT2 requires significantly more bandwidth than its predecessors to achieve its high speed.
Here is the comparison of the bandwidths:
FT8: 50 Hz
FT4: 83 Hz
FT2: ~150 Hz
FT2 occupies approximately three times the bandwidth of FT8 and nearly double that of FT4.
This increased bandwidth is a necessary trade-off to fit the data payload into a 3.8-second cycle. The developers explicitly note that because it uses ~150 Hz per signal, FT2 is not recommended for “narrow bands with limited spectrum space”.
ased on the project credits and technical documentation, artificial intelligence served as a key “development tool” for the creation of FT2.
The specific AI model used was Claude (Anthropic).
According to the sources, the AI assisted the developer (IU8LMC) in modifying the WSJT-X source code (specifically version 3.0.0-rc1),. This assistance likely focused on implementing the necessary changes to the protocol’s timing and logic to achieve the sub-four-second transmission cycle,
Yes, FT2 requires significantly better clock synchronization than FT8.
According to the technical comparison table in the documentation, the tolerance for clock accuracy is four times tighter for the new mode:
• FT8: Allows a clock error of ±200 ms.
• FT2: Requires a clock accuracy of ±50 ms.
Because the entire transmission/reception cycle in FT2 is compressed into just 3.8 seconds (compared to 15 seconds for FT8), there is much less margin for error in timing. If your computer’s clock is off by more than a fraction of a second, you may miss the decoding window entirely
