Speech to Text

How to use it

1. Pick a recognition language (English US or UK, Chinese, Japanese).
2. Click Start listening and allow the microphone permission when the browser asks.
3. Speak. Interim words appear in gray; finalized sentences accumulate in the text area.
4. Click Stop when done, edit the text freely, then Copy or Download it as a plain text file.

Common use cases

• Hands-free drafting — dictate a first version while pacing or sketching.
• Meeting and interview notes — capture the spoken words, then clean them up.
• Language practice — see your pronunciation rendered back as text.
• Accessibility — type without a keyboard for short messages and notes.

Why it stays private

The transcript is assembled and stored only in this page — nothing is saved to a server by this tool. Be aware that Chrome's recognition engine sends the audio itself to its own servers for processing, so treat dictation of confidential material accordingly. Browsers without the Web Speech API (Firefox, Safari) are told honestly that the feature is unavailable.

FAQ

Q: Which browsers support this? A: Chrome and Edge implement the Web Speech API. Firefox and Safari do not, and the tool says so instead of failing silently.

Q: Why does recognition pause after a silence? A: The engine finalizes an utterance on silence. While listening is on, the tool restarts recognition automatically and everything already recognized stays in the text area.

Q: Can I edit the transcript? A: Yes. The text area is fully editable, and Copy and Download always use its current contents.

Q: Does it need an internet connection? A: Yes for Chrome's engine, which processes audio on its servers. Without a connection, recognition will report an error.