This unicode detector uses the case sensitivity of custom block arguments. Thanks to @davidtheplatformtest for making the (value of argument (variable)) hacked block. Calling the encoder custom block is slow, because Scratch makes a new dictionary for the letters every time, but when it's already running, it can encode many letters very fast. The hacked definition block is 0.55 MB. The "unicode chars" variable is 0.18 MB. Drawbacks: - uses hacked blocks - the hacked blocks break easily - prevents TurboWarp from compiling the thread in which it's used My previous Unicode detector (without hacked blocks): https://scratch.mit.edu/projects/1035741525/