Read text aloud with highlighting in Fastreader
Read your text aloud with highlighting: choose an AI or device voice, adjust speech rate, pause and resume, and understand the current limits.
To read text aloud with highlighting in Fastreader, add your material, choose a starting point, and press “Listen”. You can use AI narration through the server or a local voice on your device. The sources differ in text handling, availability, and highlighting precision. Start with a short note that is easy to check against the original.
How to read your own text aloud
Send the meeting agenda to participants on Thursday. Before sending it, check the start time and the call link. If the time changes, tell participants in a separate message.
- Select “Add text”, paste the example, and save it with “Open text”.
- In “Text” mode, select your starting sentence and the action to read from it. Check the current position.
- Open narration settings. Check “Text language” and “Narration source”. On a phone, voice settings are in the “Voice” section.
- Press “Listen”. On a phone, find it in the “Mode” dialog. AI is selected by default; sending passages for the first time requires separate consent.
- After listening, name the deadline and both checks. Compare your answer with the note: sound and highlighting help you follow your place, but do not assess comprehension.
AI narration or a device voice?
AI narration sends passages selected for speech to our server and an external service. Read the consent dialog and select “Allow and prepare audio” only if this method suits your text. Analytics consent does not grant narration consent. Temporary audio stays on our server for up to 24 hours. To revoke permission, use “Withdraw consent and delete audio”.
For another method, manually select “Device voice” in settings. Fastreader uses only voices that the browser identifies as local; text is not sent to our server or an external narration service. You need an installed Russian or English voice. Availability depends on your device and browser, and a voice available elsewhere may be unavailable inside Telegram.
Adjust speech rate and understand highlighting
“Speech rate” ranges from 0.5 to 2. Start at 1 and slow down if you lose the meaning. The words-per-minute setting controls visual word display, not the voice while narration is enabled. Changing the speed of audio that is already prepared does not request new speech generation.
AI audio highlights the current passage as a whole because the recording has no confirmed word boundaries. A device voice highlights individual words only when it reports reliable word-start events; otherwise it also highlights the whole passage. This difference does not by itself mean that playback or your position has been lost.
Pause and resume from your place
Use the pause button. On a computer, Space also works when focus is outside fields and buttons. Look at the highlighted passage before continuing. Without exact word boundaries, part of a phrase may play again. Speech does not start automatically when you return to the page: press “Continue” when you are ready.
Why AI may not read a whole long article
The current AI feature processes passages of up to 500 characters. Project limits allow up to 1,000 new characters per day per browser or Telegram user, and up to 2,000 shared across everyone, with day boundaries in Moscow time. There is also a network limit. These are upper limits: availability additionally depends on the external service’s free allowance. Narration of a complete long article is not guaranteed.
If preparation fails or a limit is reached, the source does not switch automatically. Select “Listen with a device voice” or continue reading normally. If you open the reader in an external browser, add the original text again or transfer a library backup: the browser and mini-app keep separate libraries.
Try this technique with your own text.
Add your text →