Toolkit Dictation pad: speak, then edit
Dictation pad: speak, then edit
Get your words down by speaking, then tidy them by hand.
Everything you type or save here stays in this browser on this device. Neuro-Common does not receive it. Clearing your browser data removes it, so copy or print anything you want to keep. Back up or delete your saved tool data.
How to use it
- Read Where your voice goes. Pick your Language for speech recognition, then choose Start dictation. When your browser asks about the microphone, choose Allow.
- Speak. Words appear in Your text when the browser is sure of them. The line Hearing now shows words that may still change. Click or tap in the text first if you want new words to go in the middle.
- Turn Say punctuation out loud on to say marks such as "comma" or "new line" (English only). Turn Capitalize the first letter of sentences on or off. Keep listening after a pause starts listening again until you choose Stop dictation.
- Fix mistakes by typing in the text box. Use Undo last dictation to remove the last words you dictated, Read it back to hear your text, and Copy, Download as text or Print to use it.
- If your browser has no speech recognition, or you would rather not send your voice online, see Other ways to dictate further down this page.
Good to know
- This tool does not recognize speech itself. It uses the speech recognition that your browser offers. If your browser has no speech recognition, there is no Start dictation button, and you can still type, listen back, copy, download and print. Browser support differs, so look for the button to see what yours does.
- By default, speech recognition on a web page sends your voice over the internet to an online service, so it does not work offline. Your browser chooses that service. Neuro-Common does not receive or keep your voice or your text.
- Speech recognition can make mistakes. Read your text back before you use it. The tool shows what the browser heard. It cannot tell you whether that is right.
- Which languages work depends on your browser, and a web page cannot find out in advance which ones your browser supports. The list here is only a list of language codes to try. If one does not work, you will see a message.
- Spoken punctuation and capitalizing are simple rules written for this tool. Spoken punctuation understands only the English words listed, and while it is on it always turns those words into marks, so turn it off if you need the word itself. Capitalizing happens after a period, question mark, exclamation mark or new line, so it may also capitalize after an abbreviation that ends in a period. For Chinese and Japanese, no spaces are added between phrases.
- Dictated words go in at the cursor, and they replace any text you have selected. Undo last dictation works one step at a time, and only while you have not changed the text by hand since. After you type, use your keyboard undo or edit the text yourself.
- The pad keeps up to 60,000 characters. It is saved in this browser on this device only. Clearing the site data removes it, so use Copy or Download to keep a spare copy. Your browser may underline spelling mistakes in the box. That comes from the browser, not from this tool.
- Read it back uses the voice and speed set in the Listen controls on this page. The pad stops listening before it reads, so the microphone does not pick up the voice.
- The notes under Other ways to dictate repeat what the makers' help pages said when we read them in October 2026. Those pages change, so follow the links under Sources for your version.
Where this comes from
- This guide says that by default, speech recognition on a web page involves a server-based recognition engine, that your audio is sent to a web service so it does not work offline, and that setting the language explicitly is the recommended practice: MDN Web Docs: Using the Web Speech API
- The error codes a browser can report (such as no-speech, not-allowed, audio-capture and network), and the statement that a page cannot find out from front-end code which languages a browser supports for speech recognition: MDN Web Docs: SpeechRecognitionErrorEvent, error property
- A browser support table. When we checked, it listed no browser with full support, partial support in Chrome and Safari, and Firefox as disabled by default: Can I use: Speech Recognition API
- Microsoft's page on voice typing in Windows (the keys to press, the need for a microphone and an internet connection, and online speech recognition): Microsoft Support: Use voice typing to talk instead of type on your PC
- Apple's guide to Dictation on a Mac (the macOS 14 version): how to turn it on and start it, on-device processing on Apple silicon, and its limits: Apple Support: Dictate messages and documents on Mac
- Apple's guide for iPhone, with a version for each iOS version: Apple Support: Dictate text on iPhone
- The iOS 12 version of that guide, which lists spoken commands and says dictation may not be available in all languages or in all regions: Apple Support: Dictate text on iPhone (iOS 12 version)
- Google's page on typing with your voice on Android with Gboard, including the spoken punctuation it lists: Google Gboard Help: Type with your voice
- Google's page on allowing or blocking the microphone for a site in Chrome, and on changing a blocked site later: Google Chrome Help: Use your camera and microphone in Chrome
- The dysgraphia guide
- The dyslexia guide
- The dyspraxia guide
More tools
General support only. These tools are not medical, legal or diagnostic advice, and they do not replace professional help. If you are in crisis, go to Crisis help.