Speech to Text
Dictate with your voice and see the transcript appear in real time. Powered by your browser's Web Speech API.
ℹ️ Privacy note: Speech recognition in Chrome/Edge sends your audio to the browser's cloud recognition service (Google/Microsoft). If your audio contains sensitive information, use a local Whisper model instead. This tool itself has no server — no data is stored on our side.
Click the microphone to start
Allow microphone access when prompted by your browser.
0 words·0 chars
💡 Tips for best results
- • Use a quiet environment — background noise hurts accuracy.
- • Speak at a natural, steady pace. Over-articulating can hurt more than help.
- • Use a decent microphone — laptop mics work but headphones with a boom mic work better.
- • Chrome and Edge are the most reliable browsers for this API.
- • Interim text (gray) is a guess — wait for black final text before relying on it.
- • If recognition stops after a pause, click the mic again to resume.
🔒 Note: Speech recognition in Chrome/Edge sends audio to a cloud service (Google/Microsoft) for processing. No data is stored by CodeKeet — we have no server for this tool.
About Speech to Text
Speech to Text (STT) converts spoken words into written text. Browsers include the Web Speech API, which provides real-time speech recognition in Chrome, Edge, and (partially) Safari. Unlike server-based transcription services (which charge per minute and require API keys), the browser API is free and immediate. It works by streaming audio to a recognition service (Chrome uses Google's service), which returns live transcripts. This tool gives you a simple interface: pick a language, click the microphone, and speak — your words appear in real time. You can edit the transcript, copy it, or download it as a text file. Note: audio is processed by the browser's chosen service — for full privacy, use a local Whisper model instead.
How It Works
- Choose your language from the dropdown.
- Click the microphone button to start listening.
- Speak clearly — the transcript appears live.
- Interim (in-progress) words appear in gray; final words are confirmed in black.
- Click Stop to end, edit the transcript if needed, then copy or download.
✨ Key Features
- ✓Uses browser's built-in Web Speech API
- ✓Real-time live transcription
- ✓Interim results (gray) vs final results (black)
- ✓30+ languages supported
- ✓Continuous mode (keeps listening until you stop)
- ✓Editable transcript (type to fix errors)
- ✓Copy transcript to clipboard
- ✓Download as .txt file
- ✓Word and character counter
- ✓Recording status indicator
- ✓100% browser-based — no signup, no API key
🎯 Common Use Cases
- →Dictating emails, notes, and documents
- →Transcribing meetings or interviews
- →Accessibility — typing without a keyboard
- →Language learning — checking your pronunciation
- →Hands-free note-taking while cooking or exercising
- →Writing first drafts by speaking
- →Transcribing voice memos and ideas
- →Testing browser speech recognition for app development
💡 Advanced Tips & Pro Insights
- ▸Chrome and Edge have the best support for Web Speech Recognition. Safari has partial support. Firefox support is limited.
- ▸Speech recognition works best with a quiet environment and a good microphone. Background noise reduces accuracy.
- ▸Speak at a natural pace — not too fast, not too slow. Over-articulating can hurt accuracy.
- ▸The browser sends audio to its recognition service (Chrome → Google). For privacy-sensitive content, consider a local Whisper model (e.g., whisper.cpp).
- ▸Interim results (shown in gray) are guesses that may change. Wait for the black text before treating it as final.
- ▸If recognition stops after a pause, click the mic again — some browsers stop on silence automatically.
- ▸On mobile browsers (Chrome Android, Safari iOS), recognition works but drains battery faster. Use a shorter session for long dictation.
Frequently Asked Questions
Is this Speech to Text tool free?▾
Yes, this tool is completely free with no signup, email, or usage limits. It uses your browser's built-in speech recognition.
Is my voice private?▾
The Web Speech API sends audio to the browser's recognition service (Chrome → Google, Edge → Microsoft, Safari → Apple). It is not fully private, but it is not stored by this tool either — we have no server. For maximum privacy, use a local Whisper model.
Which browsers support Speech to Text?▾
Chrome (desktop and Android) and Edge have the best support. Safari has partial support (iOS 14.5+, macOS). Firefox support is limited. For best results, use Chrome.
What languages are supported?▾
The Web Speech API supports 30+ languages including English (US/UK/India), Hindi, Spanish, French, German, Chinese, Japanese, Korean, Arabic, Portuguese, Russian, and more. Support varies by browser.
Why does it stop after a pause?▾
Some browsers stop recognition after a silence. Click the mic button again to resume. Chrome with continuous mode usually keeps listening longer.
Does it work on mobile?▾
Yes, on Chrome Android and Safari iOS. Speech recognition works but drains battery faster than on desktop. For long dictation, desktop is recommended.
Can I transcribe an audio file?▾
Not directly. The Web Speech API transcribes live microphone input only. To transcribe an existing audio file, use a dedicated STT service or a local Whisper model.
Does it require an internet connection?▾
Yes — the recognition service (Google/Microsoft/Apple) is cloud-based in Chrome and Edge. Safari on newer iOS uses on-device recognition for some languages, which works offline. Firefox's local speech recognition (when available) can also work offline.