One question kept coming back in our engineering standups: getting a wake-word model to ~100 KB is technically possible, but which trade-offs are you willing to accept?
Wake-word models normally sit at 500 KB to 5 MB. Getting to ~100 KB required accepting three specific trade-offs that we underestimated going in.
Vocabulary is tight. A 100 KB model can hold one wake-phrase well. It cannot generalize to arbitrary phrases the way a bigger model can. If you want the phrase to change per user, or if you want it to work in multiple languages, 100 KB starts to hurt.
VoxRT Wake-Word is a ~100 KB always-on wake-word detector that runs fully on-device on iOS, Android, Linux, and in the browser via WASM. The default "Hey Assistant" model holds at 5.3% RTF sustained on a $15 Raspberry Pi Zero 2 W, and 0.16% RTF in Chrome via WASM SIMD128. Free for the standard phrase across all platforms. Custom brand phrase or language is available on a paid tier.
