Mindsmith can generate AI narration for a course or use audio files you upload. Narration can help keep learners engaged and make content accessible to more people.

You can narrate an entire lesson at once or configure the audio for individual tiles.

If your lesson is fully drafted, you can generate narration for the entire lesson in a few clicks. This is the fastest way to add audio across its pages.

To narrate your entire lesson:

A loading screen will appear while your audio generates. For larger lessons, this may take a few minutes to complete.

The voice picker shows the voices available for the lesson language and your plan. Paid plans include additional voice choices. If a selected voice does not support the lesson language, Mindsmith displays a warning before generation.

You can customize how narration behaves for specific layouts or individual tiles. This is helpful if you want to skip audio for certain sections or rewrite the spoken script without changing the on-screen text.

To adjust these settings, select a page and open Narration, then use the Page tab. From here, you can choose between three main modes:

Tiles mode narrates each tile in the layout individually using its visible text content. This is the standard mode for most layouts.

For Experience (branching scenario) tiles, the tab label changes to All Nodes, and narration is generated across all nodes in the scenario.

Use Manual mode when the spoken audio should differ from the text on the screen, such as when adding context or a conversational aside.

To generate manual narration:

If you need to discard changes to the manual script, click Sync Text to restore the visible text from the tile. A single tile cannot exceed 4096 characters of narration.

If you have visual layouts, title screens, or activities that do not require audio, select the None tab. This disables narration entirely for the selected layout or tile.

Interactive Narration plays audio in sync with how learners interact with a tile: rather than narrating the entire tile at once, each individual item gets its own clip that plays when the learner reaches it. For example, each flashcard card narrates as it flips, each process step narrates as it advances, and each accordion panel narrates when opened.

Interactive Narration is experimental and only appears when it is enabled for your workspace. Turn it on while generating lesson-wide narration. A lesson using Interactive Narration cannot mix in the other narration modes.

You can upload a recording of your own voice or audio from a voiceover artist.

To upload custom narration:

Mindsmith supports MP3, WAV, M4A, and OGG audio. After upload, enter alternative text for the audio if needed. You can remove the file and return to generated narration at any time.

You can configure narration to play automatically when a learner reaches a new section, so they don't have to click play manually.

To enable autoplay:

When enabled, audio narration will begin playing as soon as the learner navigates to a narrated section. Learners can still pause and control playback using the audio player.

Note: Most web browsers block audio autoplay until the user has interacted with the page (clicked, tapped, or pressed a key). This is a browser-level policy, not a Mindsmith limitation. Autoplay will work after the learner's first interaction with the lesson. In incognito or private browsing mode, browsers may enforce this policy more strictly.

When using Manual mode, use line breaks and punctuation to influence the pacing of the generated narration. Results may vary by voice.

On a paid organization, an Admin can open Settings and Members → Usage → Narration model and choose the default ElevenLabs model for the teamspace:

The setting applies when a compatible ElevenLabs voice is used. ElevenLabs v3 narration is limited to 5,000 characters and uses IPA entries from the organization’s Narration Glossary. If you need v3 pronunciation control, add and preview the glossary entry before generating the narration.

Mindsmith's AI voices support multiple languages:

Enterprise organization Admins can use Settings → Narration Glossary to control how specific words are pronounced. Narration glossaries are separate from translation glossaries.

Create separate entries for each language that needs pronunciation guidance. Saved rules are used when generating narration for that organization.

When learners take your course, they have full control over how they consume the narration. A sleek audio player appears at the bottom of the content area for narrated tiles.