TurboWarp Web Speech

Speech recognition and speech synthesis in the browser for TurboWarp projects, with no API key.

Install

Load turbowarp-web-speech.js as a custom extension with Run without sandbox enabled. Use Chrome, Edge, or Safari for speech recognition.

Listen and answer

when green flag clicked
set recognition language to [ja-JP]
set speech language to [ja-JP]
say (listen and wait)
speak (join [you said ] (recognized text)) and wait

Let the user interrupt

when someone starts speaking
stop speaking

Good to know

• Chrome and Edge normally send microphone audio to a cloud service for recognition. Tell users before you use it.
• TurboWarp Desktop can speak, but usually cannot recognize speech.
• Pressing the stop button releases the microphone and silences speech.
• If something fails, check last speech error.

Full block reference: README.