No. Text-to-speech conversion uses the browser's own Web Speech API (`speechSynthesis`), the same engine that powers screen readers and accessibility apps on the operating system. The text never leaves your device, even when the selected voice is a "network" voice (some system voices rely on an online service from the browser or OS vendor to synthesize the audio, but that is the browser's own decision, not this tool's, and the voice list already flags which ones are local and which depend on the network).