Audio Overview
Turn your sources into a conversation between two hosts, or a talk by one narrator, written by a chat model and spoken with ElevenLabs voices.
An Audio Overview is a conversation about your sources between two hosts, or a talk by one narrator, like a short podcast. A chat model you choose writes the script from the selected sources, and ElevenLabs speaks it with the voices you pick.
Before you start
Audio Overviews need an ElevenLabs API key. Open Settings → AI providers → ElevenLabs and paste your key; it is used only for speech, so your chat model stays as it is.
Options
| Option | Choices |
|---|---|
| Speakers | Two hosts, or one narrator |
| Format | Deep dive, brief or debate (both sides, for a narrator) |
| Length | Short (about 3 minutes), medium (about 7) or long (about 12) |
| Script written by | Any chat model you have set up; the prompt bar's by default |
| Voices | From your ElevenLabs account; their names are the hosts' names |
| Voice model | eleven_v3 or eleven_v4, which speak the hosts as one dialogue |
Each voice has a play button in the dialog, so you can hear a sample before you choose. The ElevenLabs key is sent with this request only.
Listen
The player shows a wave for each host that moves while they speak, the recording's shape as a seek bar, and buttons to skip 15 seconds or change the speed. The transcript beside it reads like a conversation and follows along; choose any line to jump there, and point at a citation mark to see the passage it draws on.
The recording is kept on your device with the notebook, and included in backups.
Download
Download the recording as MP3, or the transcript as Markdown.
The Studio
better-vault's Studio turns your sources into reports, flashcards, quizzes, data tables, mind maps, slide decks, infographics, Anatomy figures and whiteboards.
Report
Write a briefing doc, study guide, blog post or a format of your own from your sources in better-vault, with citations, a table of contents and Word export.