Before you start
PhraseLens requires macOS 26.0 or later. Download the speech model for your caption language and grant only the permissions needed by your chosen audio source.
PhraseLens — Immersive Captions
Quick answers for setup, captions, and learner-chosen lookup sites.
PhraseLens requires macOS 26.0 or later. Download the speech model for your caption language and grant only the permissions needed by your chosen audio source.
Open System Settings → Privacy & Security → Screen & System Audio Recording, enable PhraseLens, then relaunch the app and refresh sources.
Check Microphone and Speech Recognition permissions, then confirm that the selected input is receiving audio.
Confirm the recognition language matches the content, the speech model is downloaded, and the source audio is clear.
Try Reload or Open in System Browser. Third-party sites can change; you can select another current site in PhraseLens.
Try Search Link first when the result URL contains your sample phrase. Use Record Search when the URL does not contain it or does not change after searching.
Eric Li
A tested domain-based support email will appear here before release. Include your PhraseLens version, macOS version, and a short description. Do not send private audio or captions unless you choose to.