You Can Now Generate Podcast-Style 'Audio Overviews' Directly in Gemini AI

Last year, Google released a niche feature in its experimental NotebookLM product that turned out to be a sleeper hit. You could upload any document to NotebookLM, and it would use AI to generate a two-person conversational podcast out of it. These aren't meant to be uploaded online for viewership, but for you to listen to as a learning ad.

The idea is that there are times when it's easier to understand a concept when two people are just casually conversing about it. If that sounds appealing to you, good news: you no longer need to dive into the unfamiliar NotebookLM interface to try it out. Google's AI-generated podcasts, called Audio Overviews, are now available for free directly in the Gemini app and website. And now that Audio Overviews are integrated directly into Gemini, you can even use Gemini's own Deep Research reports as sources for your podcast.

Personally, I've found it useful to first prompt Gemini to create a Deep Research report on a topic, then directly generate an Audio Overview from it (skipping past having to read it myself).

Generate and download podcasts on any topic

To get started, use the Gemini website or the Gemini app. To upload your own document or slide show, click the Plus button and add in your desired source file. As soon as the file is processed, you'll see a button for Generate Audio Overview.

Uploading files and generating audio overviews.

Credit: Khamosh Pathak

Press the button and wait for Gemini to begin its work. It can take anywhere between 3–5 minutes to generate your podcast audio, depending on the depth of the material.

Alternatively, you can prompt Gemini to create a report for you using the Deep Research option below the text box. Here, too, you'll have to wait a couple of minutes until the research is ready. Then, open the Deep Research document, click the down-arrow button, and then use the Generate Audio Overview button. You can also just enter "Generate Audio Overview" in the text box.

Generate Audio Overview from Deep Research report.

Credit: Khamosh Pathak

When the processing is done, you'll get a notification from the Gemini website or the app. You'll now see a player in the chat box. Hit the Play button to start playback, and use the seek bar to jump to any point in the recording. Yes, there's speed control too, but you can only increase the playback speed up to 1.5x.

Podcast style Audio Overviews player in Gemini.

Credit: Khamosh Pathak

I got 10 minutes of podcast overview from a 12-page document, so the detail isn't bad. If you don't want to listen to your podcast right away, or if you'd like to share it, you can also download the audio for offline payback. Click the three-dot menu button in the audio player and choose the Download option.