A teleprompter, audio and video recording in one place. Read your full script or speak more freely with talking-point cards.
Write it. Choose a mode. Say it.
Bring your own writing
Write or paste your script on the home page. Open prompter takes you to the reading screen. Try a sample lets you explore without replacing your own writing.
Tap Start
Choose one of the four modes from the bottom bar. Your choice starts speech following and, in recording modes, capture together. There is no separate microphone or recording start.
Allow access, then speak
Every mode uses a microphone; video modes also ask for a camera. Wait for the microphone-ready message. A red indicator and timer show when a recording is running.
Four modes for your ideas.
All four follow your own voice. Prompter only does not create an audio or video file to play back.
01
Prompter only for reading aloud
Speech following · no recording
For presentations, public speaking and speeches. The script follows your pace; you can also turn pages manually.
Example: put the key points of a talk on cards, place your phone in front of you and speak in your own words. The prompter helps you keep your place.
02
Prompter + audio recording
Speech following + saved audio
For podcasts, monologues, poetry, interviews and learning a script by listening back. Record a voiceover for another video editor, or make a faceless video in SubiDub.
Example: record a short story, then choose a background image and subtitles in Recordings. SubiDub can turn the audio into an MP4 video, or export just the audio as M4A.
03
Prompter + selfie camera
Speech following + front camera and audio
Make video content using a full script or talking-point cards. The cards help you keep your train of thought while looking back into the camera.
Example: outline a personal introduction on three cards. Hold the phone at eye level, glance at the next thought, then speak to the camera.
04
Prompter + rear camera
Speech following + rear camera and audio
For promotional videos, product demonstrations, narration over cutaways and interviews. Film what you are showing while reading your script or interview questions on the screen.
Example: point the rear camera at a product and explain its benefits using the prompter. The video and your voice are saved together.
Adjust things while you read.
Text size with a live preview
The bottom slider and A− / A+ buttons immediately resize your actual text. Choose a size that lets you see the next lines, too. Your reading position is preserved.
Full script or talking-point cards
Use Settings at the top right to switch views. Full text suits precise reading; Outline helps you speak from short reminders. Edit card titles and reminders through Menu → Edit cards on the home page.
Pause, resume and stop
Pause appears only while following or recording is active. It pauses both; Resume starts them again. Before starting, use the arrows to turn pages manually. Pauses are excluded from the saved media. Stop ends following and finalizes the recording. Use the arrows to turn pages manually.
When the marker feels too late or too early
In Settings, increase Words ahead if the marker lags, or reduce it if it moves too soon. These display settings do not change the timing of exported subtitles.
Full camera preview behind the script
Camera modes show the complete recorded frame behind the prompter. Portrait or landscape framing is fixed when recording starts. The selfie preview is mirrored; the saved image is not. Prompter text and controls are not burned into the original recording.
From recording to a finished file.
1. Recording: after Stop, play back your saved recording. The original file is also available separately. Reopen this workspace from the library icon or the home-page menu.
Export audio as M4A or an MP4 with a background image; export camera recordings as MP4. “My written script on the video” is the single switch for including manuscript captions. “Save edits” saves edited rows; it does not create a new video.
3. Download: after Create video or Create audio file, this step shows processing status. Review the finished preview and use the download button. Change settings to create another version; earlier files remain available.
Caption characters always come from the manuscript saved with the recording, preserving punctuation and capitalization. Speech-following word timestamps provide timing. Display look-ahead, manual page turns and highlight delay do not change export timing.
The timed-word count shows coverage. Skips, improvisation, uncertain matches, connection failures or the following time limit can leave gaps. Missing timing is not invented and recognition errors do not replace the script. Request separate audio processing to recover timing, including for older recordings, or edit manually. Manual edits are protected.
Android: bubbles, overlays and camera permission
If Chrome asks you to close bubbles or overlays, close floating chats, screen filters, screen-recorder controls or floating sidebars first. Android overlays from other apps can obstruct permission prompts. SubiDub’s glass-style web controls are not Android overlays.
In Android Settings → Apps → Chrome → Permissions, allow camera and microphone access. Names vary by device. If necessary, disable the interfering app under Special app access → Display over other apps.
Then use Chrome → ⋮ → Settings → Site settings → Camera and Microphone. If subidub.com is blocked, allow it. If it is not listed yet, return to the page and choose your recording mode after closing the overlay. You do not need to deliberately deny permission repeatedly.
Each start makes one device request, without automatic retries. A website cannot change Android permissions. If access still fails, close other camera apps and restart Chrome.
Speech following currently runs for up to 60 seconds per start. Audio or video recording can continue with manual page turns. Restart following reconnects to the same recording, subject to your available allowance.
Each recording is limited to 10 minutes or 256 MB. Moving the page to the background stops capture and saves what has been recorded. Check its status after a call or screen lock.
Automatic following needs an internet connection. ElevenLabs recognizes the audio; when needed, OpenAI also receives the script and a short recognized text excerpt to find your place. Prompter only still uses online speech processing, but does not create a saved recording.
If uploading fails, Recordings offers retrying or downloading the local copy. Download important work: clearing browser data can remove local copies and guest access.
Does Prompter only read the text out loud for me?
No: you speak, and the prompter follows. AI read-aloud is a separate feature under Menu → Read aloud on the home page. It is not one of these four start modes.