Introduction to AI-Powered Podcast Features
Google’s NotebookLM has gained attention for its innovative features, particularly the Audio Overviews. This feature offers a podcast-style discussion that can be created with just one click. It mimics a conversation between a male and female host, providing engaging content that includes witty banter and insightful discussions.
While there are several tools that replicate other aspects of NotebookLM, such as Mind Maps or source-grounded summaries, none have managed to match the quality of Audio Overviews. I tested some of the top competitors but was largely disappointed. However, another Google Labs experiment called Illuminate comes close to addressing my concerns with NotebookLM's Audio Overviews.
Recently, Speechify, known for its text-to-speech capabilities, introduced its own AI podcast feature. I was eager to test it and see if it could surpass NotebookLM in this area.
What Does Speechify’s AI Podcast Feature Offer?
Speechify launched its AI Podcasts feature on August 12th as part of its back-to-school 2025 release. The feature is described as one of the company's biggest new launches, intended to provide users with a new way to consume content. Similar to NotebookLM's Audio Overviews, Speechify allows users to convert any article or document into a "lecture-style" podcast featuring 2–3 speakers discussing key topics.
The podcasts typically range from five to ten minutes long, and users can choose from different style formats. Currently, the feature is available to all Speechify Premium users on iOS. Although I wasn’t subscribed to the premium plan, I was able to test the feature through a free trial. Despite being a paid feature, I wanted to see if the experience was significantly better than NotebookLM’s free offering.
How Does the Feature Perform?
To test the feature, I used the default Podcast format and uploaded a PDF of my recent article about VS Code extensions. The process was enjoyable, with playful visuals and background music. It reminded me of ElevenLabs' podcast feature. After about a minute, the podcast started playing, and a small text indicator showed the progress in real-time.
Unlike NotebookLM, where you often have to wait minutes before hearing anything, Speechify allowed me to start listening while the rest processed in the background. The podcast generated by Speechify was 6 minutes and 6 seconds long, hosted by a man and a woman. The result was impressive, with natural jokes and clever analogies. The AI voices were among the most natural-sounding in the industry.
Speechify also included real-time captions that synced perfectly with the dialogue, making it easier to follow along. Additionally, the ability to save AI Podcasts directly in the app was a convenient feature.
Where Speechify Falls Short
Despite its strengths, Speechify's AI Podcast feature still has room for improvement. While NotebookLM's Audio Overviews don't provide an in-depth overview, they allow for longer podcasts, which is beneficial for study purposes. Speechify caps podcasts at ten minutes, limiting flexibility for detailed research.
Moreover, Speechify only allows generating a podcast from one document at a time, which isn't ideal for synthesizing information from multiple sources. The feature is currently limited to iOS, which is inconvenient for users who primarily use laptops. Finally, the paywall is a drawback, especially when free alternatives offer similar functionality.
Conclusion
Speechify's AI Podcast feature is promising and likely the closest I've come to matching NotebookLM's Audio Overviews. I'm excited to see how it evolves and whether it might eventually replace NotebookLM for me.
