Turn any idea into a finished video — powered by Claude
Enter a topic or paste a script. AI writes it, narrates it in your language, adds visuals and captions, and hands you a ready-to-post video — from a 30-second clip to a 10-minute explainer. No editing, no timeline, no fuss.
How to make an AI video in three steps
No timelines. No editing software. No learning curve. Just tell it what you want. ⚡ One-time setup: connect your Pollinations account — every video renders on it, so you only ever pay your own usage.
Enter a topic or script
Type a subject — "how photosynthesis works" — or paste your own script. That's the whole brief.
Pick your look
Choose the language, length, and whether to use free stock footage, free stock images, AI images, or full AI video.
Get a finished video
It writes, narrates, adds visuals and word-timed captions — then you download or auto-publish.
A whole video team, automated
Scriptwriter, voice artist, editor and publisher — in one link.
AI scripts & narration
Turns a topic into a tight, professional script and a natural-sounding voiceover, automatically.
14 languages
Generate the same video in English, French, Spanish, German and more — native narration each time.
AI or real stock visuals
Every scene gets a matching AI image or real stock footage — sharp, on-topic, never generic.
Scene-by-scene editing
Every video comes back as separate scenes. Drop the ones you don't want and it re-cuts in seconds — free, and you can put a scene back.
Word-timed captions
Burned-in captions that land exactly on the words — perfect for silent, scroll-stopping social feeds.
Schedules & series
Queue videos daily or weekly and auto-publish to your YouTube channel — or plan a multi-part series and get one episode per run, each building on the last.
API & Claude connector
Generate videos from your own app, Make/Zapier, or right inside Claude with the MCP connector.
AI video examples: stock footage, AI stills, hybrid and full AI
Stock footage, AI stills, an AI-video hybrid, or full AI video — every mode below was made with this app, on any subject.
Every video comes back scene by scene
Watch the whole thing, then drop any scene you don't want. Nothing is regenerated, so a re-cut takes seconds and costs nothing — and you can put a scene back.
Or plan a whole series and walk away
Set the number of parts and a cadence. One episode is made per run, and each one knows it's “part 3 of 6”, so they build on each other instead of repeating the introduction. The schedule stops itself after the last part.
Real production value, generated automatically
Every video is composed scene by scene — with motion, music, captions and branding — so it looks made, not templated.
AI avatarsPremium
A lifelike AI presenter delivers your script on camera — pick a face and a voice, no filming needed.
Clone your voice & facePremium
Star in your own videos — clone your voice and likeness once, then front every video automatically.
Up to 10-minute videos
You choose the length — a 30-second social clip or a full 10-minute explainer, rendered on your Pollen.
Cinematic AI video
Generate full AI motion clips per scene, or mix AI video with stills for a slick hybrid cut.
Ken Burns motion
Still images come alive with smooth pans and zooms, timed to the narration — never static.
Background music
A fitting soundtrack is mixed under the voiceover and ducked so every word stays crystal clear.
Any format
Landscape 16:9 for YouTube, or vertical 9:16 for Reels, Shorts and TikTok — sized right, no cropping.
Branded intro & outro
Open and close with your name, colours and a call-to-action card so every video is unmistakably yours.
Scene-matched visuals
Each shot is chosen to match what its line is actually about — literal, on-topic, never generic filler.
Natural voiceover
Warm, human-sounding narration with the right pace and pauses — in the language you choose.
Topic or full script
Let the AI write it from a one-line topic, or paste your exact script and it films it line by line.
Publish or download
Post straight to YouTube, grab a clean 1080p MP4, or pull a no-end-card cut for anywhere else.
What normally takes twenty hours
Making a short AI film by hand is not one job, it is six. This does all of them, in about five minutes, and most of it costs nothing.
The usual way
A three-to-five minute AI film means roughly 120–150 generations for about 30 usable shots, because a third to a half never match. Then colour correction and assembly by hand. Practitioners budget 20–40 hours and around $105–150.
This way
You describe it in a sentence. The script is written scene by scene, the shots are rendered, the film is graded and cut, and you edit it by talking. About five minutes, and on real stock footage, no generation cost at all.
What is automated
Shot grammar — wide, medium, close, moving, rotated so no two scenes match. One colour grade across every scene, so stock and generated footage read as one film. Credits for every contributor, collected as it goes.
Editing stays free: dropping a scene rebuilds the film from clips already made, in seconds. Films longer than three minutes are rendered in segments and stitched together — also reusing what exists, so joining costs nothing.
Being straight about the limits: faces are not consistent between shots, there is one narrator rather than a cast, and no tool yet holds a story together for ninety minutes. We automate the craft that works around those — we have not solved them, and neither has anyone else. How the craft works →
Trailers, short films and children’s stories
A film is written as a script, scene by scene — which is a conversation, not a form. Connect the app to Claude and ask for one. It asks what you want, writes the script, and renders it here.
A trailer
15 to 180 seconds, narrated, with real shot language. The form this app is best at — a 90-second cinematic cut is a trailer.
A children’s story
Storybook illustration, a warm narrator, and a character built to stay recognisable across every scene. 45 seconds to three minutes, by age.
A longer film
Written in chapters, rendered in segments, then stitched into one film. Joining re-uses clips already made, so it costs nothing and takes minutes.
How to make a particular kind of video
Each guide opens with a video made by this app — including the ones that show the real interface.
Common questions
Is Automated Video App free to use?
Videos built from free stock footage or free stock images cost nothing and are unlimited — there is no subscription. AI-generated images and AI motion run on your own connected Pollinations account and draw down your own balance, so you are never billed by us for generation.
Do I need my own API key?
For AI visuals, yes. Generation runs on your own Pollinations (Pollen) account, which you connect once. That means you control the spend and the account owner is never charged for someone else’s videos. Stock footage and stock images need no key at all.
Can I make vertical videos for TikTok, Reels and YouTube Shorts?
Yes. Choose the frame before generating: 16:9 landscape for YouTube and websites, 9:16 vertical for Shorts, TikTok and Instagram or Facebook Reels, or 1:1 square for feed posts. Captions sit higher in vertical so the platform’s own buttons do not cover them.
Can I edit the video after it is generated?
Yes, scene by scene. Every scene is kept separately, so removing one is a re-stitch of clips that already exist: it takes seconds, costs nothing, regenerates nothing, and you can put a scene back afterwards. You can also swap any scene’s picture for one of your own.
What languages and voices are available?
Narration is available in 14 languages — English, French, Spanish, German, Italian, Portuguese, Russian, Dutch, Polish, Arabic, Hindi, Japanese, Chinese and Korean — with six English voices including British male and female. All of them are free.
Can I use it from Claude?
Yes. Add Automated Video App as a connector in Claude, then ask in chat: "make me a 60-second video about our new opening hours." Claude checks which options you want first, then creates the video, and can list and remove scenes or re-cut it afterwards.
How long can a video be?
From a 30-second social clip up to a 10-minute explainer. Length is a target rather than a hard cap: scene timing follows the measured narration, so a 30-second request with a long script commonly lands nearer 40 seconds.
Can I use my own images and logo?
Yes. Upload your own photos to use as a scene’s picture or as the first frame of AI motion, and set a brand logo that appears on every video. You can also set a daily limit on AI generations, which falls back to free stock once reached.
Your next video is one sentence away
Type an idea and watch it become a narrated, captioned, ready-to-publish video — free to try.
Create your first video