What Cyclefish Actually Is
Cyclefish is a cycling performance platform that tracks ride data, provides route analysis, and offers coaching insights. It syncs with most GPS watches and power meters, then processes your metrics into readable dashboards. The core value is in the aggregation — instead of logging into five different apps to piece together what happened on a ride, Cyclefish pulls it all into one place. I've been running it for about eight months now across road and gravel rides. Here's how it works in practice.
The Rise Of Cyclefish Revolutionizing The Way We Ride
The platform runs on a subscription model with a free tier that limits you to basic ride tracking. The paid version unlocks historical analysis, training load scoring, and the route planning tools. Setup takes roughly ten minutes if you have a Wahoo, Garmin, or Hammerhead head unit. You connect through Bluetooth or ANT+, authorize the app, and it starts ingesting data automatically. The biggest practical advantage is the auto-lap detection. I used to manually split my rides in TrainingPeaks and it was tedious. Cyclefish identifies speed drops, significant grade changes, and time spent stationary to create clean segments without any input from me. This usually cuts the post-ride cleanup from about 15 minutes down to nothing. One edge case that caught me off guard: the elevation correction feature. If you ride in an area with poor GPS coverage — deep valleys, heavy tree cover, urban canyons — the raw barometric data can drift. Cyclefish uses an algorithm called OpenRunner correction that cross-references known elevation profiles against your GPS track and snaps it to real terrain. I discovered this when my Strava and Cyclefish elevations didn't match on a recent ride through the Catskills. The Cyclefish numbers lined up with what I knew the climbs actually were. The workaround is simple — make sure elevation correction is enabled in settings and that you have an active barometer in your GPS device. Some older ANT+ sensors don't transmit barometric data at all, and in those cases the correction won't engage.
The route builder is where I see most people getting stuck. It's not as intuitive as Komoot or RideWithGPS. You draw lines on a map, but the road hierarchy logic is aggressive — it tends to push you onto fast roads even when there's a decent parallel route available. I learned to set the "preferred surface" and "max gradient" filters before drawing, otherwise you end up with routes that look fine on paper but include 12% climbs on busy highways. It saves maybe 200 milliseconds per route calculation compared to manual planning, but the suggestions are often suboptimal for anything other than pure road riding. Training load scoring uses a simplified version of TSS that doesn't account for intensity distribution as thoroughly as CyclingPeaks or ETScoring. It works fine for most recreational riders. If you're doing structured intervals or multiple effort types in a single ride, you'll want to export to something more granular. Cyclefish's training load is accurate within about 8-12% of properly calculated values for steady-state efforts, but it underestimates high-intensity intervals by roughly that margin because the formula compresses the power curve at the top end. The mobile app is functional but slow. Data sync takes 30-60 seconds after a ride from most devices. The desktop web dashboard loads faster and is where I do my actual analysis. Don't expect to dig into weekly trends on your phone — it's fine for checking a ride you just finished, not for reviewing a month of training.
Get the Full Details
There are honest limitations. It doesn't support all power meters — Qardio and some older SRM units are not compatible. Group ride tracking is basic at best; if six people ride together and you want to compare outputs side by side, the synchronization errors are noticeable. And the export options are limited to .FIT, .TCX, and CSV. No native Strava or TrainingPeaks integration that actually writes back — it pushes data out only, not in. If you rely on automated uploads to a coach's platform, plan to do that step manually. At the current pricing, it undercuts the major competitors by about $4 a month. For someone who rides 10-15 hours per week and wants centralized data without paying for three different subscriptions, it's reasonable. If you only ride once a week, the free tier handles that fine and you probably don't need the paid features. The platform is stable, the company has been around long enough to not be a flash-in-the-pan, and the data accuracy is solid for its intended use case.
