How to use Speech to Text
Dictate text using your microphone and see it transcribed in real time — entirely in your browser via the Web Speech API's SpeechRecognition interface, with no audio ever transmitted to any server. Click Record and start speaking; words appear as you talk, with punctuation inserted automatically. Ideal for hands-free note-taking, voice drafting, accessibility workflows, and generating first drafts significantly faster than typing. Supported natively in Chrome and Edge on desktop and Android.
- Click "Start Recording" and grant microphone permission when the browser prompts you.
- Speak clearly at a measured pace — transcribed text appears in the output area in real time.
- Pause briefly between sentences to allow the recognition engine to finalize each phrase.
- Click "Stop Recording" when you are finished speaking.
- Review and edit the transcript, then copy it to the clipboard or export it as a plain-text file.
Your data never leaves your device — 100% private processing.
How the Web Speech API SpeechRecognition works
The Web Speech API's SpeechRecognition interface connects to the browser's built-in speech recognition engine. In Chrome and Chromium-based browsers, recognition is handled by Google's speech infrastructure. The API delivers two result types: interim results (partial transcripts that may still change as the engine processes more audio) and final results (committed text that will not change). This tool uses continuous mode with interimResults enabled, so live text appears as you speak before each phrase is finalized — giving immediate feedback without waiting for silence.
Tips for better transcription accuracy
Accuracy depends on microphone quality and ambient noise. A USB headset or cardioid microphone in a quiet room produces the best results. Speak at a measured pace — very fast speech increases error rates noticeably. For technical or domain-specific terms, speak more slowly and edit afterward. The Web Speech API supports dozens of languages via BCP-47 locale codes (e.g., "en-US", "fr-FR", "zh-CN"); selecting the correct language for your speech significantly improves recognition accuracy. Background music or overlapping voices are the most common cause of transcription errors.
Glossary
- Web Speech API
- A browser-native JavaScript API comprising SpeechRecognition (speech-to-text) and SpeechSynthesis (text-to-speech) interfaces, available without installing any external libraries.
- SpeechRecognition
- The Web Speech API interface that captures microphone audio and converts it to text using the browser's recognition engine, streaming results in real time.
- Interim result
- A provisional transcription segment returned by SpeechRecognition that may still change as the engine processes more audio context before the phrase is finalized.
- Final result
- A committed transcription segment from SpeechRecognition that will not change, typically produced after the speaker pauses or reaches the end of a phrase.
- BCP-47
- An IETF standard for language tags (e.g., "en-US", "es-MX", "zh-CN") used to specify a locale for speech recognition or synthesis engines.
Related reading
Frequently Asked Questions
Why use Speech to Text?
- Transcribe spoken words into editable text in real time with no audio file uploads
- Generate first drafts two to three times faster than typing for most speakers
- Supports dozens of BCP-47 locale codes for accurate recognition across multiple languages
- Interim results display live so you can monitor transcription quality while still speaking
Common use cases
- Dictate a rough blog post draft while stepping away from the keyboard
- Capture verbal meeting notes directly into a text area for immediate editing and sharing
- Transcribe recorded interview responses by playing audio through the microphone input
- Convert voice memos into editable text for a conference presentation outline
- Practice a foreign-language speech and verify pronunciation by checking the transcribed output
Get weekly tool tips & updates
New tools, power-user tips, and productivity hacks — delivered free every Friday.
No spam, ever. Unsubscribe with one click.
Related AI & Writing
Text to Speech
Convert text to natural-sounding speech online for free. Choose a voice, adjust rate and pitch, and listen instantly. Private text-to-speech in your browser.
Online Notepad
A free online notepad that autosaves in your browser. Write notes, drafts, and to-do lists with live word count and no sign-up. Private and offline-friendly.
Readability Score Checker
Check the readability of your text online. Get Flesch Reading Ease, Flesch-Kincaid grade level, and word, sentence, and syllable counts instantly.
Text Summarizer
Summarize long text online with extractive sentence ranking. Generate concise summaries from articles, notes, and reports directly in your browser.
Text Similarity Checker
Compare two texts for similarity online. Get a percentage match using cosine and Jaccard similarity to detect duplicate or near-duplicate content.
Headline Analyzer & Generator
Analyze and generate headlines online. Score titles for length, emotion, and power words, then create click-worthy variants for blogs, ads, and emails.
Explore all AI & Writing.