Upload audio, transcribe it via the OpenAI Speech-to-Text API, then delete words to cut them out of the audio. Your API key stays in your browser and audio is sent only to OpenAI.