One question that comes up in the 67 Challenge community regularly: does the platform matter? iPhone vs Android — does it affect your score?
Here’s the honest breakdown.
How the detection works on both platforms
The 67 Challenge uses Google’s MediaPipe Hands model on both platforms. This means the underlying AI is identical — the same model, the same hand landmark detection logic, the same wave-counting algorithm.
What differs between platforms:
- Camera hardware — front camera quality, frame rate, and sensor size vary
- Processing power — how fast the AI model runs per frame
- OS-level camera APIs — how fast the app can get each camera frame
iPhone performance
On modern iPhones (iPhone 12 and later), the Neural Engine chip runs the MediaPipe model at a consistent 30+ fps with very low latency. The front camera on recent iPhones produces sharp, low-noise images even in lower light — which helps hand detection stay accurate.
Best iPhone results come from: iPhone 12 or newer, good lighting, face-centered in front of the camera.
Android performance
Android has a wider range of hardware. On high-end Android phones (Pixel 7+, Samsung Galaxy S22+), performance is comparable to iPhone — fast inference, high frame rate, consistent detection.
On mid-range devices (3-4 year old phones), you may notice:
- Slight lag in detection
- Higher chance of missed reps at very high wave speeds
- Slower recovery if a hand briefly leaves frame
Budget Android phones may have a lower front camera frame rate (as low as 15fps), which limits how many waves can be accurately counted per second.
Does it change your score?
In practice, good lighting matters more than your phone model. A mid-range Android in bright daylight will outperform a flagship iPhone in a dim room.
If you’re chasing Legendary, the limiting factor is almost never the device — it’s technique and lighting. See the hand position guide and high score tips for what actually moves the needle.
The browser version: platform differences
On the browser version, iOS Safari has improved significantly with iOS 16+. On iOS 15 and below, WebAssembly performance can limit the model to 15–20fps, which does affect detection at high wave speeds. If you’re on older iOS, download the app instead for better performance.
Chrome on Android handles WebAssembly well across most mid-range and above hardware.
Verdict
For pure performance: modern iPhone or high-end Android are equal. For budget devices: iPhone tends to be more consistent. But for most players at most scores, the platform is not the bottleneck.
Download free on your platform of choice:
Or play in your browser right now.