Changelog

Closer voices, clearer numbers, and word follow-along

You can now choose how closely the hosts match their voices, how strongly to colour the delivery, how numbers are read, and how far they sound. New episodes highlight each spoken word as you listen.

• Match the chosen voice more loosely or more closely, and exaggerate its character.
• Always read numbers and dates in full, or keep digits as written.
• Move the hosts a little farther from the microphone, without adding a room echo.
• On a new episode, click a word in the player to jump to that moment.

Spoken delivery controls and clearer speech

You can now set how fast and how steadily the hosts speak, and keep names and acronyms pronounced the way you want. The player has a new clearer-speech option.

• Slow down or speed up the spoken delivery, and pick a livelier or a steadier voice.
• Lock the spoken language on the Flash and Expressive voice models, so a host stays in the language of the episode.
• Rewrite names and acronyms for speech before the audio is made.
• On the Expressive voice model, keep audio tags such as laughs and sighs in the script.
• The clearer-speech option runs in your browser. The audio is not sent anywhere.

Buy a plan before you create an account

The pricing page now takes you straight to checkout, without a sign-in first. The plan goes to the email address you enter at checkout.

• A new address gets a new account, and you are signed in right after the payment.
• If the address already has an account, the plan goes to that account. Sign in with the email link to use it.

Chapters, playback controls and conversation options

New episodes have chapters, speed and skip controls, and a transcript you can click to jump to a line. In the generator you can name the hosts, choose who speaks first, pick a conversation style and add a pause between turns.

• Chapters of about 90 seconds each, as a list you can click and as a WebVTT chapters file.
• Playback speed from 0.75× to 2×, and buttons that skip 15 seconds back or forward.
• Four conversation styles: balanced, interview, debate and explainer.
• A pause of 0.25, 0.5 or 0.8 seconds between turns, also for a script that you write yourself.

SRT and WebVTT subtitles

Download subtitles for new episodes as SRT or WebVTT files, timed from each spoken turn.

• The subtitle downloads are on the episode page, next to the audio and the transcript.
• Episodes made before this release keep the plain-text transcript download.

A new studio with ElevenLabs voices

AIdeaFlow Podcast has a new design, and every episode is now voiced by ElevenLabs with three voice models: Flash, Multilingual and Expressive.

• The site is available in 12 languages, with right-to-left layouts for Hebrew and Arabic.
• Each plan card shows the episode length, the voices and the monthly audio of the plan.

Your podcasts: a public switch and downloads

Your podcast library now shows 24 episodes per page, and each episode has a switch that makes it public or private.

• Episode pages offer the audio and the transcript as downloads.
• The audio download names its real format: MP3 for new episodes, WAV for some older ones.
• When you delete an episode, its audio and transcript files are deleted too.

Episodes finish after you close the tab

Generation now runs on our servers until the episode is saved, so you can close the browser tab and find the episode in Your podcasts later.