AIToolBase

Tool Guides · 7 min read

AI Podcast Generator: How to Make a Podcast with AI from Your Notes or PDFs

Two practical routes, a one-click study podcast with Gemini Notebook (formerly NotebookLM) or your own script read by an AI voice, plus the rules to follow before you publish.

By AIToolBase Editorial Team · Published

Gemini Notebook banner on a white background with the logo, the headline "Think Smarter, Not Harder" with "Smarter" in a green-to-blue gradient, and a black "Try Gemini Notebook" button

Short answer: The easiest free AI podcast generator is Google's Gemini Notebook (the new name for NotebookLM): upload your notes or PDFs, choose Audio Overview in the Studio panel, and it creates a conversation between AI hosts that you can download. For a podcast you publish under your own name, write a script with ChatGPT or Claude, voice it with a text-to-speech tool such as ElevenLabs, then edit and upload it. Either way, tell listeners the voices are AI, and only use material and voices you have the rights to.

This guide is for students who want to listen to their study material, and for creators curious about AI-voiced shows.

Which route is right for you?

RouteBest forCost (as of November 2026)Okay to publish?
Gemini Notebook Audio OverviewRevising notes, papers and PDFsFree with a Google account, within usage limitsGoogle's help doesn't cover publishing; see the rules
Gemini Notebook Video OverviewA narrated explainer video of your sourcesCounts toward your Gemini Notebook usage limits; check your planSame as above
Your script + ElevenLabs FreeTrying out AI voicesFree, 10,000 credits a monthNon-commercial only, with credit in the title
Your script + paid AI voiceA real show you want to grow or monetiseElevenLabs Starter is $6 a monthYes, Starter includes a commercial license

A quick reality check: an Audio Overview is an AI summary of your sources read by AI voices. Google says Audio Overviews "may contain inaccuracies or audio glitches". In our view, it is a great study and listening aid, not a replacement for reading or for a show with your own ideas.

How do you make an AI podcast with NotebookLM (Gemini Notebook)?

Google renamed NotebookLM to Gemini Notebook in July 2026. It's the same product, and notebooklm.google.com now redirects to notebook.google.com.

  1. Go to notebook.google.com and create a new notebook.
  2. Upload your notes, PDFs or other files as sources.
  3. In the Studio panel, select Audio Overview.
  4. Pick a format: Deep Dive, The Brief, The Critique or The Debate.
  5. Choose a language and a length.
  6. Optionally, type a focus prompt, such as "Explain chapter 3 for a beginner".
  7. Start generation and keep working; it runs in the background.
  8. Play it, then select Download to save the audio file.

What the four formats do, according to Google:

  • Deep Dive (default): two hosts unpack and connect the topics in your sources.
  • The Brief: one speaker covers the key takeaways in under two minutes.
  • The Critique: two hosts give constructive feedback on material like an essay.
  • The Debate: two hosts argue different perspectives in a formal back-and-forth.

Length options are Shorter, Default and Longer, and Longer is English only. Audio Overviews come in 80+ languages, defaulting to your Google Account language.

This short video from MIM Technovate (September 2026) shows the whole flow with several economics PDFs: uploading, choosing Deep Dive, adding a focus prompt and downloading.

How to Turn PDFs into a Podcast with NotebookLM — MIM Technovate

Did you know? You can interrupt the AI hosts. Turn on Interactive mode when you create an overview, select Join while listening, and ask a question by voice; the hosts answer from your sources, then carry on. Google says it is English only, and your voice isn't stored.

Free limits. As of November 2026, Gemini Notebook uses compute-based limits rather than a fixed number of overviews. Your quota refreshes every 5 hours until you reach a weekly limit; Google AI Plus doubles it and AI Pro quadruples it. Check Settings → Usage to see what's left. Free accounts can add up to 50 sources per notebook.

Sharing. Besides downloading, you can share a link, but the notebook must be shared with the listener or set to Anyone with a link. Deleting the overview breaks the link.

Google's original 2024 demo below shows the idea in under a minute. The interface and the product name have changed since, so use the steps above.

Use NotebookLM to generate audio discussions — Google

How do you make a podcast video with AI?

Gemini Notebook can also turn your sources into a narrated video.

  1. In the Studio panel, select Video Overview.
  2. Pick a format: Explainer, Short (about 60 seconds) or Cinematic.
  3. Choose the narration language and, if offered, a visual style.
  4. Select Generate now or Generate later.
  5. When it's ready, use the download icon to save the video.

Google warns generation can take more than 30 minutes. Cinematic is English only and limited to users 18 and over, and some options may be missing in the mobile app.

How do you make your own podcast with an AI voice?

This route gives you full control over what's said. Write first, then voice.

  1. Paste your notes into ChatGPT or Claude with the prompt below.
  2. Read the script aloud once and fix facts and stiff sentences.
  3. Paste the script into ElevenLabs' text-to-speech tool.
  4. Choose a voice from its library and generate the audio.
  5. Download the file and trim it in a free editor such as Audacity.
  6. Add a spoken line saying the voice is AI-generated.
  7. Export an MP3 and upload it to your podcast host.
Turn these notes into a 5-minute solo podcast script.
Audience: [who]. Tone: friendly and clear, like explaining to a smart friend.
Structure: a 2-sentence hook, 3 main points with one example each, a 1-line recap.
Write for the ear: short sentences, no bullet points, no headings.
Use only facts from my notes and mark anything you're unsure about with [CHECK].
Notes: [paste here]

ElevenLabs free plan terms. As of November 2026, the Free plan includes 10,000 credits a month, and text to speech uses about one credit per character. ElevenLabs says free-plan audio "cannot be used for any commercial purpose", and if you publish it you must put "elevenlabs.io" or "11.ai" in the title. Paid plans include a commercial license, and instant voice cloning starts on the paid Starter plan.

Prefer to compare voices? Our speech APIs category lists other text-to-speech tools. If you record your own voice instead, a free speech-to-text tool can turn the episode into a transcript for show notes.

What are the rules for publishing an AI podcast?

Four things to get right before you hit publish:

  • Rights to the material. Apple Podcasts' guidelines say creators must "only include content that they have the rights to use". Turning someone else's book or paid course into a public episode can break copyright, even if AI did the reading.
  • Disclosure. Apple's guidelines require creators who use AI to generate audio, including synthetic voices, to "prominently disclose this to audiences in the content and metadata" of each episode and show. Our recommendation: say it in the audio and the description on every platform.
  • No impersonation. Spotify said in May 2026 that it will remove shows that copy other creators' voices "using AI voice cloning or any other method".
  • Gemini Notebook output. Google's Audio Overview help page doesn't address publishing. In our view, keep overviews for personal study unless you have checked Google's terms and your sources' copyright.

Never clone a real person's voice without their clear, written permission, including friends, teachers or celebrities. Our guide to AI voice cloning for voiceovers covers consent and safe use.

Want to listen to a PDF but also get written notes? Our guide on how to summarize a PDF with AI pairs well with an Audio Overview.

Frequently Asked Questions

Is there a free AI podcast generator?

Yes. Gemini Notebook (formerly NotebookLM) makes Audio Overviews free with a Google account, within usage limits that refresh every 5 hours.

How do I make an AI podcast from a PDF?

Upload the PDF as a source in Gemini Notebook, select Audio Overview in the Studio panel, choose a format, and download the result.

Is an AI podcast good for studying?

It's a useful way to review material on a walk or commute. Google warns overviews may contain inaccuracies, so check key facts against your sources.

How do I make a podcast video with AI?

In Gemini Notebook, select Video Overview in the Studio panel and choose Explainer, Short or Cinematic. Generation can take more than 30 minutes.

How do I create a podcast using an AI voice?

Write a script with ChatGPT or Claude, paste it into a text-to-speech tool such as ElevenLabs, then edit and publish the audio. Free ElevenLabs audio is non-commercial only.

Can I publish an AI podcast on Apple Podcasts or Spotify?

Yes, if you own the rights to the content and disclose AI voices; Apple requires that disclosure. Spotify says it will remove shows that impersonate other creators with voice cloning.

Sources

Reviewed on November 3, 2026. Google renamed NotebookLM recently and changes Audio and Video Overview options and usage limits often, and ElevenLabs updates its plans regularly, so check the official pages before you rely on a detail.