SayHere Support

Get started

1. Open SayHere and allow Microphone and Speech Recognition. Accessibility and Input Monitoring access are not required.
2. Check the dictation language on Prepare speech models. Initial preparation may need an internet connection to download models. Installation and analyzer initialization are verified before you can continue.
3. Each recording uses one selected language; SayHere does not automatically switch based on what you say. To test another language, choose it on the model page or under Language in Settings and wait for its models. The interface language is not the dictation language.
4. The default shortcut is Control plus the grave accent key, without Shift. Press once, release the keys and speak, then press again to stop. Do not hold it down. On a U.S. keyboard this key is below Esc and left of 1. The floating button also starts and stops recording. If you customized the shortcut, use the combination shown in Settings.
5. After stopping, wait for Copied, click the intended field, and press Command-V. During onboarding, paste into the test field and then click Done. No fixed phrase is required; speak in the selected language.

Models are still preparing

Permission does not mean the models are installed. Actual download progress appears when Apple provides a total; otherwise the page shows waiting. Verification continues after 100%. Keep a network connection and free storage available. You can retry after failure or timeout, or cancel preparation. Changing language prepares that language’s models.

Recording does not start or returns no text

Confirm SayHere is running and check Microphone and Speech Recognition permissions. If the shortcut conflicts, use the recording button and change the shortcut in Settings. Speak toward the Mac’s built-in microphone. If a recording is protected or recognition fails, open pending recordings in History and retry the matching entry. You can still start another dictation. Text is copied, so paste it manually into the intended app.

Local data

History, vocabulary, and protected recordings stay on your Mac and can be deleted in the app. Apple’s system services manage the initial model download. Dictation uses on-device recognition without automatic cloud fallback.