Speech to Text
Speak into your microphone and get live transcription. Supports multiple languages via the browser Web Speech API.
What is the Speech to Text?
The Speech to Text tool listens to your microphone and streams a live transcription onto the screen using the browser Web Speech API — the same technology behind Chrome dictation. It supports 10+ languages including English variants, Hindi, Spanish, French, German, Portuguese, Japanese and Chinese.
Interim words can appear while you speak and are replaced as the recognizer gains context; finalized phrases remain available as editable text that you can correct and copy. Selecting the matching language and speaking in a quiet space usually improves recognition of names and punctuation.
ToolSink never receives or stores your audio or transcript. Note that on Chrome, the browser itself sends audio to Google's recognizer service to compute the transcription. This distinction matters for confidential conversations: local page handling does not mean the browser’s recognition engine is necessarily offline.
How to use the Speech to Text
- 1Pick your language from the dropdown.
- 2Click "Start listening" and grant microphone permission when prompted.
- 3Speak clearly — the transcript appears in real time.
- 4Click "Stop" to lock the final text, then copy it with one click.
When to use the Speech to Text
- Draft notes or an email by dictating when typing on a phone is slow.
- Capture a rough transcript of your own spoken outline before editing it into prose.
- Practise pronunciation and inspect which words the selected language model recognizes.
- Create accessible text from a short voice memo played clearly near the microphone.
Key features
- Live interim and final transcription
- Supports 10+ languages
- Editable transcript with copy-to-clipboard
- Works with any wired or Bluetooth microphone
Important notes
- Chrome and other Chromium browsers offer the smoothest experience; Safari has partial support.
- On Chrome, audio is processed by the browser's built-in recognizer (Google's cloud service). ToolSink itself never receives or stores your audio or transcript.
Frequently asked questions
Why do words appear and then change?
The engine emits interim results as it hears more speech and finalises them once the sentence completes.
Can I dictate for a long time?
Yes, but some browsers auto-stop after a few minutes of silence — just click Start again.
Is recognition completely offline?
Not necessarily. ToolSink does not receive the recording, but Chrome commonly sends audio to Google’s speech-recognition service. Browser and platform behavior can differ.
How can I improve transcription accuracy?
Choose the correct language variant, use a close microphone, reduce background noise, speak in complete phrases and manually correct specialized names or technical vocabulary.
All Utility
Tap any tool to jump straight in — no signup, works on mobile.