Type a topic and this AI podcast generator writes a back-and-forth conversation between two AI hosts, gives each one a voice and a face, and renders it as a captioned vertical video podcast clip. Clips run from 15 seconds to 3 minutes, sized for TikTok, Reels and YouTube Shorts.
The podcast format lives in the FacelessGenie creator. Your topic, hosts and editing style are set there, and you see the credit cost before it renders.
An AI video podcast clip in the dynamic cuts style: each speaker full screen, hard cuts on every turn, captions on each line.
The same format in the mixed shots style, adding two-shot moments between the cuts so both hosts are seen reacting.
The classic split layout keeps the host and guest on screen together for the whole clip, like a clipped interview podcast.
Describe what the two hosts should argue about, for example "is it smarter to rent or buy in 2026" or "the strangest rules on long-haul flights". You can also paste a script you wrote.
Keep the default host and guest or upload a photo for each one. Choose dynamic cuts, mixed shots or classic split screen, and pick a length between 15 seconds and 3 minutes.
The AI writes the dialogue, voices both speakers, lip-syncs them, adds captions and exports a vertical MP4 you can download and post.
The output is an MP4 video. One AI host opens with a sharp question or claim, the guest answers with a specific point, and the two trade turns until the guest lands the last line. Each speaker is lip-synced to their own voice, captions follow every line, and the whole clip renders as an MP4 in 9:16.
Tools such as NotebookLM's Audio Overviews turn your notes into a long audio episode that you listen to. This page is for the other job: a short podcast-style clip that people watch in a feed. If you only need the sound from a clip you made here, run the MP4 through the free extract audio tool and you get an MP3.
Short two-person debates do well as Shorts and Reels because the cut between speakers gives the eye a reason to stay. Podcasters use them as teasers for a full episode, faceless channels use them as a recurring series format, and educators use them to turn one dense idea into a quick question-and-answer exchange.
On a phone, a podcast clip has to hook a viewer in its opening line. In this format the first line is a challenge rather than a greeting, each turn is three to seven seconds long, and the speakers alternate every beat so the frame keeps changing. Captions sit on screen for people watching with the sound off. Because the hosts are generated, you can post a new debate every day without booking a guest or setting up a microphone.
Dynamic cuts shows one speaker full screen at a time and cuts hard on each turn, which suits fast debates. Mixed shots adds two-shot moments between the cuts so viewers see both hosts react. Classic split keeps both speakers on screen the whole time, the look most people know from clipped interview podcasts. The examples on this page show all three, rendered from the same format.
Most channels that use this format run it as a series: the same two hosts, a new question each episode. Keep the host photos and voices the same across clips and viewers start to recognize the pair. If you already run an audio show, the clips work as trailers for full episodes, and the free podcast name generator helps if the show still needs a name.
Creating a FacelessGenie account is free, and new accounts get trial credits. Each video podcast clip uses credits based on its length and the avatar quality you pick, and the exact cost shows before you render.
Not directly. This format renders a video clip with two on-screen hosts. You can pull the audio track out of the finished MP4 with the extract audio from video tool, but for hour-long audio episodes an audio-first tool is a better fit.
Yes. Upload a photo for each speaker and the format animates it. The studio portrait option restyles each photo into the same podcast set, or you can switch it off to use the photos exactly as uploaded.
Between 15 seconds and 3 minutes, with 45 seconds as the default. Most clips that hold attention on Shorts and Reels land between 30 and 60 seconds.
The AI writes the conversation from your topic: the host asks or challenges, the guest answers with a concrete detail, and the last line is the guest's strongest point. You can review and edit the script before you render, or paste your own.
Use the free podcast name generator for that. Describe the show and it returns ten name ideas per batch, which you can check against Apple Podcasts and Spotify before you commit.