Mestrio Teleprompter Help and User Guide

Applies to version: 1.0
Last updated: August 20, 2026

Mestrio Teleprompter creates and manages scripts and displays them in the App or over another app. Scripts remain on the current device by default and are not automatically synced to the cloud.

1. Quick Start

1. On Home, select Create script.
2. Start with a blank script, resource, system file, or AI draft.
3. Review the title, text, and line alignment in the editor, then save.
4. Open the script from Home and adjust text and scrolling.
5. Press Play for automatic scrolling, or switch to Voice for speech follow.

Android asks for authorization when microphone, overlay, or legacy file access is first needed. Refusing a permission does not affect features that do not require it.

2. Home and Script Library

Home lists saved scripts and creation shortcuts. You can:

• open a script in the prompting panel;
• edit an existing script;
• delete a script you no longer need;
• create from blank text, resources, a system file, or AI; and
• open Resources to manage imported documents.

Deleting a script also deletes its saved reading position and script-specific display settings and may be irreversible. Keep an original document or separate backup for important material.

3. Create and Edit Scripts

3.1 Blank script

Choose Start with an empty script, then enter a title and text. The editor organizes formatting by line and does not add a separate paragraph-spacing model. A newline creates another line.

3.2 From Resources

Choose Choose from resources and turn imported text into an editable script. The resource and script are independent. Removing the resource later does not remove a script already created from it.

3.3 From a system file

Choose Choose a system file to use Android's picker. Current support:

The current limit is 10 MB per file. Legacy .doc, PDF, image, and password-protected documents are not supported. Complex DOCX layout, images, comments, headers, and some list styling will not be preserved completely.

3.4 Create with AI

Enter a topic, then optionally choose scenario, duration, language, tone, audience, and key points. After you select Generate:

1. The input is sent to the Mestrio AI service and Google Gemini.
2. The result opens in the regular editor and is not saved automatically.
3. Review facts, wording, rights, and suitability.
4. Save to add the draft to the local script library.

An unavailable AI entry usually means the build has no service configuration, the network is unavailable, a rate limit was reached, or the service is temporarily down. Do not submit passwords, tokens, or unnecessary sensitive information.

4. Text and Formatting

The editor supports:

• title and body editing;
• left, center, right, or justified alignment for the current line or selected lines;
• editor-preview line spacing;
• letter spacing; and
• Save or Save and Prompt.

Alignment belongs to each line. With no text selection, alignment applies to the cursor line. A selection crossing several lines applies it to every covered line. Alignment does not insert leading spaces, so it remains stable while editing.

Line spacing is extra space between text lines, not font line height. 0 sp means no extra spacing, not zero line height. Letter spacing uses em and scales with text size.

5. Prompting Panel

The top bar returns, switches scripts, or opens related setup. The bottom controls playback and common adjustments. You can:

• play or pause automatic scrolling;
• increase or decrease speed;
• change text size, line spacing, and letter spacing;
• show or hide the reading guide;
• mirror text for reflective teleprompter glass;
• switch between timed scrolling and speech follow; and
• start the floating panel from the current script.

A countdown may appear before playback. While paused, reposition the script or change display controls. The App stores reading position and display settings for each script.

6. Speech Follow

Speech follow recognizes what you read using an on-device model and advances through the script. Microphone audio is not uploaded to AI or Firebase and does not create a recording file.

6.1 First-time setup

1. Open Speech and models from My or the prompting settings.
2. Choose the language you will read.
3. If several models exist, compare Lightweight, Balanced, Accurate, and Experimental labels.
4. Download and wait for installation and verification.
5. Return to the prompting panel, select Voice, and grant microphone access.

Once installed, recognition normally runs offline. Model downloads, AI generation, and Firebase diagnostics still use a network.

6.2 Improve results

• Select a model matching the script and spoken language.
• Speak steadily and reduce background noise.
• Stay reasonably close to the script; after skipping a large section, allow time to relocate.
• Prefer a lightweight model on a low-memory device.
• Test a short section before a live session.

Waiting for speech means no usable speech was detected. Trying to find your position means text was recognized but not confidently aligned. Interrupted usually indicates a permission, audio-device, or model-loading problem.

7. Speech Models

Models are not all bundled in the APK and are downloaded by language. The App verifies each archive and does not replace a working model with an incomplete installation.

When a language has multiple models:

• Lightweight uses less download, memory, and processing;
• Balanced trades size for accuracy;
• Accurate is generally larger and needs stronger hardware; and
• Experimental may add languages or a new architecture but needs more validation.

Remove a model to reclaim storage. If it was preferred, download it again or select another installed model for the language.

8. Floating Teleprompter

The floating panel displays a script over a camera, meeting, or streaming app.

1. In the in-app prompting panel, select Show over other apps.
2. On first use, allow Mestrio Teleprompter to display over other apps in Android settings.
3. Drag the panel freely; it does not force edge snapping.
4. Use the resize handle to change width and height.
5. Open the script picker inside the overlay without first returning to the App.
6. Open language/model selection in the overlay when needed.
7. Use Return to App to reopen the current script in the in-app prompting panel.

Android shows a foreground-service notification while the panel runs. If a control is hard to tap, enlarge the panel or move it away from a system-gesture area.

9. Resources

Resources hold text extracted from imported files so it can be reused for different scripts. You can import, find, use, and delete resources.

• Content checks avoid unnecessary duplicates of the same text.
• Removing a resource deletes its extracted private copy, not the original file.
• A script created from a resource is an independent copy.
• Filenames and extracted text remain local and are not sent to Firebase.

10. Appearance and System Settings

Under My > Settings, choose Light, Dark, or System theme and whether the screen remains awake while the App is open. Languages and models use a separate management page.

Android controls overlay, microphone, and legacy storage permissions. If an in-app state differs from Android settings, the system permission takes precedence.

11. Data and Privacy

• Scripts, resources, settings, reading state, and models are mainly local.
• Clearing App data or uninstalling removes private App data and is usually irreversible.
• Speech follow processes microphone audio on device.
• AI generation sends the information you intentionally enter for generation.
• Firebase Analytics and Crashlytics are enabled by default but are not intentionally given script or recognition text.
• Read the Privacy Policy and Terms for complete details.

12. Troubleshooting

The file picker does not open or import fails

1. Update the system Files app or install a compatible file manager.
2. On Android 9 or earlier, grant storage access and use the built-in legacy browser.
3. Confirm that the file is under 10 MB and is TXT, Markdown, or DOCX.
4. Save a protected or complex document as a normal DOCX or UTF-8 TXT and retry.

The prompting panel does not scroll

• Confirm the script is not empty and Play is active.
• Check whether Voice mode is selected without a model or microphone access.
• Increase speed, or pause and reposition the reading point.

Speech follow does not advance or often relocates

• Match the language and preferred model to the script.
• Confirm that the model says it is ready.
• Re-enable microphone access in Android settings.
• Close another app that is holding the microphone.
• Reduce noise, move closer, and read text close to the script.
• On low memory, choose a lightweight model and restart the App.

A model does not download or install

• Check network and free storage.
• Keep the App open until verification completes.
• Use reinstall to fetch and validate a clean copy.
• If GitHub is inaccessible on the network, try another network later.

The floating panel does not appear

• Allow Display over other apps in Android settings.
• Confirm that the floating notification was not stopped and the App was not force-stopped.
• Some devices require disabling strict battery optimization.
• Reopen the in-app prompting panel and start floating again.

AI generation keeps failing

• Confirm network access.
• Shorten the topic or key points and retry later.
• Wait after a rate limit or service maintenance.
• If the screen says AI is not configured in this build, that APK cannot generate yet.

Newline, spacing, or alignment looks wrong

• Update to the latest version.
• Remember that 0 sp removes extra line spacing, not line height.
• Place the cursor on the line, or select all intended lines, before alignment.
• For a reproducible issue, send device model, Android version, steps, and a screenshot without sensitive script text.

13. Contact Support

Email: robinxdroid@gmail.com

Useful details in a report:

• App version, Android version, and device model;
• screen and exact steps where the issue occurs;
• whether overlay, AI, or speech follow was active;
• a screenshot or recording with private script content removed; and
• any non-sensitive error shown on screen.

Never send a password, API key, access token, identity document, or complete sensitive script.