NotebookLM Audio Overview guide: formats, languages, length, sharing, and limits
A NotebookLM Audio Overview turns your sources into an AI-hosted discussion. This guide covers the four formats, 80+ languages, length options, interactive mode, sharing rules, and the five-hour usage refresh.
A NotebookLM Audio Overview turns the sources in a notebook into a spoken discussion between AI hosts. You add sources, open the Studio panel, select Audio Overview, and optionally choose a format, language, length, and focus prompt before it generates. Two things have changed that older guides miss: Google's help center now presents these pages under the name Gemini Notebook, and usage is measured by compute with a quota that refreshes every five hours rather than as a fixed number of overviews per day.
Prepare the sources first
An Audio Overview is built from what you upload. Google describes it as an objective reflection of your source content rather than the hosts' own opinions, so the quality of the sources is the quality of the episode. Remove unrelated files and make sure the sections you care about are actually in the notebook. Studio generation shows an expected usage cost before you start, so finalizing sources first avoids paying that cost twice.
There is also a permission requirement. You need edit access to a notebook to generate or delete an Audio Overview. A colleague with view-only access cannot regenerate it for you.
Choose one of four formats
After you select Audio Overview in Studio, the generation panel lets you set preferences. Google lists four formats. Deep Dive is the default: two hosts unpack and connect the topics in your sources. The Brief has a single speaker who delivers the key takeaways in under two minutes. The Critique has two hosts give constructive feedback on material such as an essay or design doc. The Debate has two hosts hold a formal back-and-forth on the topic.
The decision rule is short. Pick Deep Dive to understand a long source, The Brief to skim it quickly, The Critique to get feedback before you submit something, and The Debate when the material benefits from opposing views.
Language, length, and the focus prompt
Audio Overviews can be generated in more than 80 languages. The default output language is the preferred language in your Google Account settings, so if you want a different language for one episode, select it in the panel instead of changing your whole account.
Length has three options: Shorter, Default, and Longer. Google marks Longer as English only. For other languages you are choosing between Shorter and Default, so if you need more coverage, split the material across notebooks or generate one episode per section.
The prompt box lets you focus on specific topics or adjust the expertise level, for example "discuss only the methods section for first-year graduate students." Later, the three-dot menu next to the episode in Studio offers "View custom prompt," which shows the prompt you used so you can reuse or refine it.
Generating and listening
Generation runs in the background, so you can create other artifacts or move to other screens while it works. Google notes it may take a couple of minutes. During playback you can change speed, rate the episode, and keep working in the notebook: explore quotes, get citations, and ask the sources questions while the audio plays.
