Quick answers to the most common questions, and a direct line if you need anything more.
Three things to know on your first run.
Pick a card from Featured, Recent, or your own Results to start a fresh episode in seconds.
While listening, tap any word in the lyrics and the narration restarts from that sentence.
While playback is active, tap the voice button to pause and pick a different voice or speed.
If your question isn't here, write to us — the email at the bottom of this page goes straight to the developer.
Yes. Every word is written and spoken on your iPhone. You can turn on airplane mode and the app keeps working exactly the same. Nothing you type, listen to, or generate ever leaves your device.
On first launch the app loads the language model into memory. After that, subsequent topics start in seconds. Generation speed depends on your iPhone — newer chips are noticeably faster.
Start a podcast, then while it is playing tap the voice button in the top-right of the player. A menu appears with five voices — Alan, Ryan, Jenny, Amy and Kathleen. Tap one to switch instantly.
Open the voice menu, then tap the voice that is currently selected (the green one). A second menu opens with five speeds: 0.75×, 1×, 1.25×, 1.5× and 2×.
Yes — tap any word in the episode text. The narration restarts from the beginning of that sentence. You can scrub freely backward or forward, just like a Spotify track.
When an episode finishes, tap Load next chapter at the bottom to continue the same thread. The model picks up using the last paragraph as context, so the new chapter flows naturally from where the previous one ended.
Home has three independent sections:
Results — topics from your last search.
Recent — the last few topics you opened.
Featured — freshly generated suggestions. Tap the refresh icon to roll a new set.
iOS automatically pauses audio when another app takes over (a phone call, navigation prompt, another music app). Tap the play button to resume. If the issue keeps happening, force-quit the app and reopen it.
Generating speech and text on-device is genuinely heavy work. On older devices (iPhone 12 mini, SE 3rd gen) the app uses a lower-power path automatically. If you still notice warmth or slowdowns, take short breaks between episodes.
iPhone running iOS 16.4 or later. iPads can install and run it in iPhone-compatibility mode (portrait window). The app is portrait-only by design.
Everything ships inside the app — the language model and all five neural voices. That is why there are no servers, no subscriptions, and no internet required. One download, forever offline.
No. No sign-in, no ads, no in-app purchases, no analytics, no tracking. The app does not connect to any server at all.
The episodes are produced by an on-device language model. Like any AI, it can make mistakes, simplify, or invent details. Treat episodes as an entertaining starting point, not as a source of fact — double-check anything important.
Write directly to the developer. Include your iPhone model and a short description of the issue and you'll get a personal reply.
sivavimelrajhen@gmail.com