The AI podcast engine

Two hosts talking.Zero recording gear

Write a tagged dialogue, pick two voices and two avatars, and Shorz animates a full interview-style podcast video — side-by-side or focused on who's speaking.

Powered by the world's best AI models

  • ElevenLabs
  • Kling
  • OmniHuman
  • OpenAI
  • Google
Two host photos and a dialogue script become a filmed two-person podcast episode

a two-person scripta few questionsfull podcast episode

Made with Shorz Podcast

A finished Shorz Podcast video: two generated hosts and a tagged script in, both lip-synced by HeyGen Avatar IV in split screen, then captions, AI B-roll, emojis and music added automatically. Press play for sound.

Podcast · 31s · 16:9 · two hosts · captions · AI B-roll · emojis · music

Script: “Host: Okay, real question. Will AI replace video editors? Guest: Honestly? It'll replace the boring part. Cutting silences, captions, resizing for every platform. Host: So editors become more like directors. Guest: Yes! One editor with AI can now ship what a whole team used to. Host: That's either terrifying or amazing. Guest: Both. Mostly amazing.”

Built for two-speaker formats

For a single presenter use Avatar. For interviews and dialogue, Podcast gives each role a voice, an avatar, and a layout that follows who's talking.

🎙️YouTube

Interview shows

Host + guest format with natural back-and-forth.

💬Shorts + Reels

Debate & opinions

Two takes on a topic, each with a distinct voice.

📰Daily series

News roundups

Anchor and analyst breaking down the week's stories.

🎓Education

Q&A lessons

Teacher answers student questions on camera.

🛍️Brand

Product chats

Founder and customer talking through the offer.

🎭Entertainment

Story dialogues

Narrative scenes with two characters trading lines.

Made with Shorz

Faceless YouTube, shorts and more. Any style, length & aspect ratio.

Layouts that follow the conversation

Tag every line, pick how both hosts appear, and optionally add camera motion and B-roll.

Writing a tagged podcast dialogue in Shorz

Tagged [Interviewer] / [Interviewee] script

Write dialogue with role tags on each line. Each speaker gets their own ElevenLabs voice and optional avatar image. Preview lines before you generate.

At least one line per role

Podcast avatar display and camera motion settings

Split view or talking focus

Show Both Avatars for a classic interview layout, or Show Only Talking Avatar to focus on who's speaking. Add Handheld Camera or Slow Zoom for motion.

Display mode + camera motion

Voices & languages that sound human

Podcast dialogue runs on ElevenLabs voices. Each host gets a unique voice, in any of 90+ languages.

90+

languages supported

1000s

voices in the library

Supported voices

One voice per host

  • Independent ElevenLabs voice per speaker — interviewer and interviewee
  • Preview lines with each host's voice before you generate
  • Audio tags for emotion & delivery — [happily], [whispers], [laughs], and more

Every voice runs on Shorz credits. No API key needed.

🌍

Supported languages

90+ languages

  • 90+ languages — narrate Avatar and Podcast scripts in any of them
  • Write the script in the target language; the voice speaks it natively
  • Ship localized variants without re-recording talent
  • Word-by-word captions stay in sync after you generate

Popular languages

Narrate Avatar and Podcast scripts in any of these and more — powered by ElevenLabs v4.

AfrikaansArabicArmenianAssameseAzerbaijaniBelarusianBengaliBosnianBulgarianCatalanCebuanoChichewaCroatianCzechDanishDutchEnglishEstonianFilipinoFinnishFrenchGalicianGeorgianGermanGreekGujaratiHausaHebrewHindiHungarianIcelandicIndonesianIrishItalianJapaneseJavaneseKannadaKazakhKirghizKoreanLatvianLingalaLithuanianLuxembourgishMacedonianMalayMalayalamMandarin ChineseMarathiNepaliNorwegianPashtoPersianPolishPortuguesePunjabiRomanianRussianSerbianSindhiSlovakSlovenianSomaliSpanishSwahiliSwedishTamilTeluguThaiTurkishUkrainianUrduVietnameseWelsh

From dialogue to video in four steps

One workflow that turns a tagged script into a two-host conversation video.

Setting interviewer and interviewee voices and avatars
1

Set up both characters

The chat asks for the host and the guest. Pick a voice and an avatar image for each. Upload a photo or generate one with AI.

Two voices · two avatars

Writing tagged podcast dialogue in Shorz
2

Write the dialogue

Tell the chat the topic and it writes the conversation. Or paste your own lines like [Interviewer] Hello… and [Interviewee] Thanks for having me…

Shorz marks who says what

Configuring podcast display mode and camera motion
3

Set camera & finishing

The video shows whoever is talking. Pick a still, handheld or slow-zoom camera, then captions, music and extra shots if you want them.

Camera · captions · optional overlays

Publishing a podcast video to YouTube and TikTok
4

Create, then publish

Pick the format, hit Create Video, review the result, and upload straight to YouTube and TikTok. Reels-ready for Instagram.

One-click upload, built in

Let an AI agent produce episodes for you

Shorz isn't a locked web app — it speaks MCP. So Claude, Cursor, Codex or Gemini can write tagged dialogue, set voices and avatars, generate the episode, and upload to YouTube and TikTok. Batch a season from a single message.

  • Draft and produce whole episode batches overnight
  • Publish to YouTube & TikTok without leaving the chat
  • You stay in control — review before anything posts
Works withClaudeCursorCodexGemini

Connected to Shorz MCP

“Write a 4-minute interview about AI startups with [Interviewer]/[Interviewee] tags, two distinct voices, show both avatars, captions on, export 16:9 for YouTube.”

Weekly interview

  • 1Draft tagged dialogue
  • 2Set voices + avatars
  • 3Configure split layout
  • 4Generate podcast video
  • 5Upload to YouTube

One dialogue in. A two-host episode out.

Why creators skip the recording studio

Real podcasts still need gear. Shorz podcast videos need a tagged script and two faces.

Traditional podcast video

  • Mics, cameras and editing for every episode
  • Finding two people who can record on the same day
  • Avatar tools that only support one speaker
  • No layout control when both hosts appear
  • Separate tools for dialogue, captions and publish
  • Subscriptions that meter every minute

Podcast in Shorz

  • Write the dialogue — Shorz animates both hosts
  • Two unique voices and avatars per episode
  • Split view or focus on who's talking
  • Optional camera motion and topical B-roll
  • Finish and upload to YouTube & TikTok in-app
  • Prepaid credits — no subscription, never expire

Pay once. no subscription

Free videos

$0

no card needed

Try it free

Your first 3 videos are free. Edit, clip, or make a faceless video.

Claim 3 free videos

Free download for Windows & Apple Silicon Mac

  • Auto Edit & Clipping3 videos shared
  • Faceless videoScript up to 90 sec
  • Source footageFull edit up to 10 min
  • Local exportsIncluded
  • PublishingNot included
  • SchedulingNot included

Starter

$10

one-time · 1,000 credits

Get started

Make your first videos and see what Shorz can do for your channel.

  • Faceless video5+ min
  • Auto-edited video10+ min
  • Short clips20+ clips
  • Video ads4.5+ ads
  • Music video1+ min
  • Avatar & podcast0.5+ min
  • Publishing100+ posts
  • SchedulingIncluded

Podcast questions

Each non-empty line must start with [Interviewer] or [Interviewee], then a space, then the spoken text. Include at least one line for each role. Tags are matched case-insensitively.

Avatars are optional but recommended. Upload a photo per role or generate portraits with GPT Image 2.6 / Nano Banana 2.1. Each still is animated and lip-synced to that speaker's voice.

Each host gets an independent ElevenLabs voice from the full library. Write the tagged dialogue in any of 90+ languages and both speakers generate in that language.

The frame follows whoever is speaking. You can add handheld or slow-zoom camera motion. An AI agent over MCP can also set a split layout that shows both hosts.

In the chat. Give it a topic and it writes the dialogue, or paste your own and it marks who says what. Type finishing cues like B-roll or music in the chat box too.

Yes. Over MCP, Claude, Cursor, Codex or Gemini can write tagged dialogue, set voices and avatars, generate the video, and upload — you review before anything goes live.

Cost scales with total dialogue length: two speakers mean TTS and avatar animation for each role. Podcast videos run on credits from the first render; packs start at $10, they never expire, and there's no subscription. (The 3 free videos cover Auto Edit, Clipping and Text-to-Video only.)

Still have questions? Ask founder