2 speakers · up to 10 min
AI video podcast generator — two hosts from one photo, two voices
Put both hosts in one photo, give each one their own voice track and let a two-speaker engine make them talk to each other. For solo shows, one host can talk for up to 10 minutes; cut the best moments into vertical clips for Shorts and Reels.
- Two hosts, one track each
- Up to 10 min per video
- Solo host engines too
- Your voices or AI voices
Updated by the Kenerate AI team
On this page
Quick answer
What is an AI video podcast generator?
An AI video podcast generator turns podcast audio into a video of the hosts talking. On Kenerate you do it in Avatar Talk: upload one photo with both hosts, add one audio track per person, and a two-speaker engine (InfiniteTalk Duo) animates each face to its own voice for up to 10 minutes. It costs 9 credits per second at 480p, or 4.5 with the Fast Duo draft engine.
Key facts
- Speakers on screen
- 2 (one photo, one track each)
- Longest two-host video
- 10 min (InfiniteTalk Duo)
- Free plan
- Yes — 25 sign-up credits
Real demo
One episode, a second language
Switch versions and turn the sound on. A dub changes the voice; a lip-sync pass in Avatar Talk redraws the mouth to match.
- Original (English). The podcast host clip as generated.
- Spanish dub. Dubbed with Kenerate's dubbing tool: new language, same lips.
- Spanish + lip sync. VEED Lipsync 2 in Avatar Talk redraws the mouth to the Spanish audio.
All three are unedited outputs of Kenerate's tools.
Lip sync a dubbed episodeReal results
Podcast-style talking clips
A solo host, a two-person interview, a show intro and a sign-off: real Wan 3 clips with generated voices, not Duo engine outputs. In Avatar Talk the voices are your own tracks.
AI-generated people. Open a clip to copy both prompts.
How to
How to make an AI video podcast
- Time needed
- Steps
- 4 steps
Step 1
Write the episode as a dialogue
Mark every line with the host who says it. Short turns, questions and reactions read better on screen than long monologues.
Step 2
Make one audio track per host
Record each host yourself, or voice each part in Voice Studio with a different voice. Line the parts up in any audio editor so each track is the full episode length, silent while the other host speaks.
Step 3
Prepare one photo with both hosts
Use a sharp image of two people side by side, both faces toward the camera — a photo of you and a co-host who agreed, or two characters you generated.
Step 4
Generate with a Duo engine
In Avatar Talk pick InfiniteTalk Duo (or Fast Duo for a cheap draft), add the left and right tracks, set "Who speaks when" to Together and generate.
See it in the app
Open Video Studio in Text to Video mode and write the shot — subject, action, camera move and light.
Watch the tour · 1:19
Text, image & reference to video, models, edit, extend & chat
Chapters
Checked 22 October 2026
What an episode costs
Worked out from Avatar Talk's per-second rates. Voices are extra if you make them in Voice Studio (15–60 credits per 1,000 characters, depending on the engine).
| Episode | Engine | Credits |
|---|---|---|
| 30 s two-host clip for Shorts | InfiniteTalk Fast Duo | 135 |
| 1 min two-host segment, 480p | InfiniteTalk Duo | 540 |
| 1 min two-host segment, 720p | InfiniteTalk Duo | 1,080 |
| 10 min two-host draft | InfiniteTalk Fast Duo | 2,700 |
| 10 min solo episode, 720p | Kenerate Avatar | 4,500 |
Engines
Engines for video podcasts
Five Avatar Talk engines that suit podcasts: three for two hosts, two for long solo shows. Prices are Kenerate credits per second of billed speech.
- 2 speakers
WaveSpeed
InfiniteTalk Duo
- Input
- Photo of two + one track each · up to 10 min
- Output
- 480p (9/s) or 720p (18/s)
Open InfiniteTalk DuoBest for
Full two-host episodes
- Cheapest
WaveSpeed
InfiniteTalk Fast Duo
- Input
- Photo of two + one track each · up to 10 min
- Output
- 360p draft · 4.5/s
Open InfiniteTalk Fast DuoBest for
Cheap drafts and test episodes
Meituan
LongCat 1.5 Duo
- Input
- Photo of two + one track each · first 64 s
- Output
- 480p (12/s) or 720p (24/s)
Open LongCat 1.5 DuoBest for
Short two-person clips and teasers
Kenerate
Kenerate Avatar
- Input
- One photo + up to 10 min · body-language prompt
- Output
- 720p (7.5/s) or 1080p (13.5/s)
Open Kenerate AvatarBest for
Solo hosts in clean 1080p
SoulX
SoulX FlashHead
- Input
- One photo + up to 10 min
- Output
- 480p (4.5/s) or 720p (9/s)
Open SoulX FlashHeadBest for
Long solo episodes on a budget
From Avatar Talk's engine list (22 October 2026). Two-speaker engines bill the longer track when hosts talk together, and both tracks added up when they take turns.
Tips
Video podcasts that look like a real conversation
Most problems come from the photo or the audio, not the engine.
Both faces need to be clearly visible, roughly the same size, with closed mouths.
Keep both tracks exactly the same length so the speakers stay in sync.
Use two clearly different voices — viewers should never wonder who is talking.
Draft on InfiniteTalk Fast Duo, then re-run the keeper on InfiniteTalk Duo at 720p.
Longer than 10 minutes? Render the episode in parts and join them in an editor.
Only animate yourself, characters you created, or people who agreed to it.
FAQ
AI video podcast generator — questions
8 questions
How do I make a two-person AI video podcast?
Put both hosts in one photo, make one audio track per host and open Avatar Talk with InfiniteTalk Duo. Add the left person's track and the right person's track, choose who speaks when, and generate. Each face moves to its own voice, for up to 10 minutes per video.
Link to this answerCan I turn my existing podcast audio into a video?
Yes. Split the recording into one track per host (most editors can export each speaker separately), keep both tracks the full episode length, and use them with a photo of the two of you. For a solo show, use one track with Kenerate Avatar or SoulX FlashHead.
Link to this answerCan the AI write the podcast script and pick the voices for me?
Not in one click. You write or paste the dialogue, then voice each host in Voice Studio with any of 555 preset voices or your own cloned voice. The AI podcast generator page explains the audio side step by step.
Link to this answerHow long can an AI video podcast be?
Up to 10 minutes per video on InfiniteTalk Duo, Fast Duo, Kenerate Avatar and SoulX FlashHead. LongCat 1.5 Duo uses only the first 64 seconds. For longer episodes, render in parts and join them in an editor.
Link to this answerHow much does an AI video podcast cost?
It depends on the engine and length: a minute of two-host video is 540 credits on InfiniteTalk Duo at 480p, 1,080 at 720p, or 270 on the Fast Duo draft engine. A 10-minute solo episode on Kenerate Avatar at 720p is 4,500 credits.
Link to this answerIs there a free AI video podcast generator?
Avatar Talk works on the free plan. The 25 sign-up credits are enough for a few seconds of a Fast Duo test, so you can check your photo and voices before buying credits for a full episode.
Link to this answerCan I make vertical podcast clips for Shorts and Reels?
Yes. Generate the clip from a vertical photo of the hosts, or cut 15–60 second moments from the full video in your editor. A 30-second two-host clip on Fast Duo is 135 credits.
Link to this answerCan I put a celebrity or someone else in my video podcast?
No. Use your own face, a co-host who agreed to it, or characters you generated. Animating real people without their consent is not allowed, and platforms such as YouTube ask you to label realistic synthetic content.
Link to this answerIn depth
AI video podcasts, explained
On this page
Why two tracks and not one mixed file
A two-speaker engine has to know which face says which words. That's why InfiniteTalk Duo takes one audio track per person: the left track drives the left face, the right track the right face. If both tracks run the whole episode with silence where the other host talks, choose "Together" and you pay for the episode length once.
Kenerate vs a one-click podcast app
Apps like HeyGen's AI Video Podcast write the dialogue from a topic, URL or PDF and pick two AI hosts for you, with paid plans from $24 a month. Kenerate is the opposite trade: you bring the script and the voices, and you see the price per second before you generate.
- Kenerate: your photo, your voices, up to 10 min, 4.5–18 credits per second, no subscription needed.
- One-click apps: topic in, episode out; less control over who appears and how they sound.
Disclosure
If the hosts look or sound like real people, say that the video is AI-generated. YouTube asks creators to label realistic altered or synthetic content, and listeners trust a show more when it is upfront about its AI hosts.
AI podcast generator
Usually means the audio: a script read by AI voices, with intro music. The AI podcast generator page covers that part, and this page turns the result into video.
Two-person talking avatar
The same Duo engines, used for interviews, sketches or sales dialogues instead of a podcast.
Explore
Every Kenerate video tool
Generators, tools, models and guides — all in one studio.
