YouTube channels
A weekly show without recording
Publish a two-host episode every week on a steady set with the same faces, and keep the channel on schedule when nobody can sit down to record.
AI podcast generator that puts two hosts on camera
Give Fliki a topic or a finished script. It casts one or two hosts, builds the set, writes the conversation, and films it like a real show: a wide of the table, singles of whoever is talking, every line lip-synced.
Standard and Premium plans · 16:9, 9:16 or 1:1 · Episodes up to 10 minutes
Why creators pick Fliki for video podcasts
An audio podcast generator stops at the voices. This one films the show.
Most AI podcast tools hand back an MP3. Fliki returns the episode as video, with hosts you picked, a set you described and a camera that cuts like an editor would.
Two hosts, one table
Both hosts sit side by side facing the camera, turned toward each other, so the wide reads as a real two-person show rather than two pasted portraits.
A directed camera
An AI director cuts between the wide and a single of whoever is speaking, and keeps the wide to under a third of the runtime so the episode never sits on one shot.
Every line lip-synced
Once the voices are recorded, each take is lip-synced on MiniMax H3 Max by default, with P-Video Avatar and Fabric as alternatives.
Your hosts or new ones
Pick a character from your library and keep the same face across every episode, or describe a host and Fliki casts one for you.
Review before you spend
The host portraits, the wide and each close-up are drawn first for you to approve or redraw, before any voice or lip-sync credits are used.
Idea or finished script
Type a topic and Fliki writes a two-voice conversation at 1, 2, 3, 5 or 10 minutes, or paste your own script and it runs to its own length.
Natural Gemini voices
Hosts speak in Gemini voices across 80+ languages, so the same show can run in English, Spanish or Hindi without re-casting.
Any set you can describe
A cosy studio with a brick wall, a rooftop at dusk, a bright morning-show desk. One sentence sets the room, and every shot is drawn inside it.
Every platform
Render in 16:9 for YouTube, 9:16 for Shorts, TikTok and Reels, or 1:1 for the feed, then edit any shot in the Fliki editor.
How it works
How to make a video podcast with AI in 4 steps
From a topic to a filmed, lip-synced episode.
Step 1
Give it a topic or script
Type what the episode is about and how long it should run, or paste a script with a label on each speaker.
Step 2
Pick hosts and a set
Choose one or two hosts from your library or let Fliki cast them, pick their voices, and describe the room they sit in.
Step 3
Approve the shots
Review the host portraits, the wide of the table and each single. Redraw anything that is off before the episode is made.
Step 4
Get the episode
Fliki records the conversation, cuts the camera to whoever is talking and lip-syncs every take. Edit it or export it.
Use cases for an AI video podcast generator
One topic. A filmed episode.
A video podcast without booking a studio, a guest or a camera operator.
Clips after
Then cut it into Shorts
Once the episode exists, the clip generator pulls the lines that stand alone and cuts them vertical with captions.
Brands
A company show on a budget
Two hosts talk through the product update, the customer story or the industry news, on a set in your brand colours.
Training
Lessons as a conversation
A host asks the questions a learner would ask and the other answers, which holds attention longer than one presenter reading slides.
AI podcast generator FAQ
Frequently asked questions about AI video podcasts
How Fliki films a podcast, casts the hosts, and compares to audio-only generators.
Tools
Discover more
AI video models
Discover models
AI video podcast generator · Standard and Premium
The show is one topic away.
Pick the hosts, describe the set, and Fliki films the conversation.
Make an episodeFree forever plan · No credit card required · Cancel anytime