How this works in your browser
This uses the browser’s Web Speech API for recognition, and it is worth being direct about what that means: in Chrome and Edge the browser streams audio to a cloud speech service rather than processing it on your device. That is how those browsers implement the feature, not something layered on here, and it is the reason this is the one tool on the site where audio genuinely leaves your machine. Recognition uses surrounding words to disambiguate, which is why natural pacing outperforms careful word-by-word speech. Support is limited to browsers that implement the API, with Firefox and Safari offering little or none.
Who uses Speech to Text
Drafting quickly
Get a rough first version down faster than typing allows.
Hands-free notes
Capture thoughts while doing something else.
Accessibility
Enter text without typing where that is difficult.
Transcribing your own recordings
Play back non-sensitive audio and capture a rough transcript.
Frequently asked questions
Is my audio uploaded to a server?
Speech recognition in most browsers, including Chrome, uses a cloud speech service under the hood as part of the browser’s built-in engine. This is a browser platform behavior, not something Slaytic Converter controls or adds on top of.
Which browsers support this?
Chrome and Edge have the most reliable support for the Web Speech API. Firefox and Safari have limited or no support at the time of writing.
Do I need to grant microphone access?
Yes, your browser will prompt you to allow microphone access before transcription can start.
Should I dictate anything confidential?
No, and this is the one tool on the site where we would say that. Because the browser routes recognition through a cloud service, spoken audio leaves your device by design. That is a property of the browser platform rather than something added here, but the practical advice is the same: do not dictate medical details, legal matters or anything else you would not send to a third party.
How do I get better accuracy?
Speak at a normal conversational pace rather than slowly and deliberately, since the recogniser uses surrounding words for context and isolated words give it nothing to work with. Reduce background noise, and use a headset microphone if you have one, which typically helps more than anything else.
How do I add punctuation?
Say it. Speaking "comma", "full stop" or "new paragraph" is recognised as the mark rather than the word in most cases. Otherwise expect an unpunctuated block that needs a pass afterwards, which is normal for dictation.
Why does it not work in my browser?
Because support is genuinely uneven. Chrome and Edge implement the speech recognition API reliably, while Firefox and Safari offer limited or no support. Unlike most features here, there is no workaround: the capability either exists in your browser or it does not.