AI Slideshow Video Generator
Pick a category or write your own script. AutoFeed narrates it, generates one image for every few seconds of audio, burns in captions timed to the syllable, and publishes the result.
- Category presets
- 27, plus a blank brief
- Images
- One per ~5s of narration, minimum 3
- Caption styles
- 12 plus none
- Languages
- 39 locales
- Credits
- 1,000 / 1,200 / 1,500 by length
What a AI Slideshow Video Actually Is
An AI slideshow is the workhorse faceless format: a written script, a synthetic narrator, and a run of generated stills that change every few seconds while word-level captions track the audio. It is the format behind most of the facts, stories and top-five channels on short-form, because it scales — the script is the only part that has to be good.
AutoFeed derives the number of images from the actual voiceover rather than from a guess. The script goes to ElevenLabs, the true length of the returned audio is divided by five and rounded up, and that many image prompts are rendered, with a floor of three. That is why the images never run out halfway through a long story or pile up at the end of a short one.
AI Slideshow Videos Made With AutoFeed
Every clip below came out of this generator with no manual editing afterwards. Press play — nothing autoplays on this page.
How the AI Slideshow Generator Works
Every step below is a screen in the editor, in the order you meet it.
- 1
Pick a category, or bring your own
The category grid holds 28 tiles. The first one is Custom Series Brief — it is the default, and it is not a topic at all: it hands you an empty script box. The other 27 are the topic presets, in grid order Italian Brainrots, Brainrot, Freaky Gram, Kids Stories and Cat Stories at the head, then Crazy Animals, Celebrity Facts, Top 5 List, Messi And Ronaldo, History Facts, Wars & Conflicts, Conspiracy Theories, True Crime, Scary Stories, Movie Reviews, Car Reviews, Luxury Lifestyle, Travel Destinations, Business Stories, Financial Tips, Productivity Hacks, Life Hacks, Tutorials, Science Simplified, New Technologies, Psychology Facts and Religion & Prayers.
- 2
Set the target length
Three targets: 15 to 30 seconds, 45 to 60, or 60 to 90. This is what the script writer aims at, and it is the biggest single lever on both retention and cost — the image count is derived from the finished voiceover, so a longer script means more images and more render time. The three lengths cost 1,000, 1,200 and 1,500 credits. 15 to 30 is the default and is what most channels should be shipping most of the time; 60 to 90 exists because TikTok’s Creator Rewards only pays on original video over a minute.
- 3
Generate the script
You get back a title, a description with hashtags and the narration script. All three are text boxes: rewrite as much as you want, or paste in a script you already have and ignore the generator entirely.
- 4
Choose the look
Seven art styles — Realistic, Cartoony, Anime, Cyberpunk, Retro, Painting and Comic — applied as a modifier to every image prompt, so the whole video holds together instead of drifting between styles shot to shot.
- 5
Choose the voice and language
Any voice from the ElevenLabs library, filterable by gender and by use case: narrative story, narration, social media, news, meditation, conversational, ASMR or animation. There are 39 language and locale options, from English (US, UK, Australia, Canada) through Japanese, Hindi, Portuguese, Spanish, Arabic, Tamil and Vietnamese.
- 6
Captions, particles and movement
Twelve caption styles plus None, and a Custom tile that opens a builder with 40 fonts, colours, outline width, glow and uppercase. Seven particle overlays — smoke, ember, dust, sakura, glowy, travel and warp — plus none, which is the default. Camera Shake adds drift to stills so the frame is never completely still; Animated Images instead converts each still into a generated clip and switches the shake off.
- 7
Render, then let it run
Export at 9:16, 16:9 or 1:1. Then download, or hang the format off a series and let AutoFeed keep publishing to TikTok and YouTube Shorts.
Everything You Can Configure
These are the controls this format actually ships with, and the values behind each one.
Category
The preset the script is written from.
- 28 tiles: 27 topic presets plus Custom Series Brief
- Italian Brainrots, Brainrot, Freaky Gram and Kids Stories lead the grid
- Custom Series Brief — the first tile and the default — for a blank script box
Duration target
What the script writer aims at, and what the video costs.
- 15 to 30 seconds (default) — 1,000 credits
- 45 to 60 seconds — 1,200 credits
- 60 to 90 seconds — 1,500 credits
Art style
Applied to every image prompt in the video.
- Realistic (default)
- Cartoony, Anime, Cyberpunk
- Retro, Painting, Comic
Voice and language
ElevenLabs narration with character-level timing.
- Full ElevenLabs voice library
- Filter by gender and by use case
- 39 language and locale options
Caption style
Word-level captions burned into the render.
- YouTuber (default), Supreme, Barbie, Neon, Glow, Thumbnail
- Glitch, Creepy, TikToker, Instagramer, Classic 1, Classic 2
- Or None
- Custom — a builder with 40 fonts, including Luckiest Guy, Titan One, Bangers, Bungee, Creepster and Poppins
Particles
A full-frame overlay on top of the images.
- None (default)
- Smoke, Ember, Dust, Sakura
- Glowy, Travel, Warp
Movement
Two mutually exclusive ways to add motion.
- Camera Shake — drift across the stills, on by default
- Animated Images — each still becomes a generated five-second clip
- Turning on Animated Images disables Camera Shake
- Animated is priced per clip: 2,400, 4,800 or 7,200 credits at the three lengths
Background music
Optional bed under the narration.
- None (default)
- 87 tracks in the shared library, shelved as Viral, Classics, Moods and Movies
- Including the 11 original beds: Solitude, Assumptions, Passo Bem Solto, Calm Viral, Epic, Focused Viral, Happy, Mysterious Viral, Relaxing, Sad, Scary
Aspect ratio
The shape of the finished file.
- 9:16 vertical
- 16:9 horizontal
- 1:1 square
24 Slideshow Prompt Templates
A slideshow is only as good as the sentence it starts from. These are templates rather than finished scripts: fill the brackets, pick the matching category preset, and let the generator write the narration around it.
Fill in the brackets and paste the result into the Topic box on the Content tab, or select the category listed, press Edit on the post and paste it into the Script box to write the whole thing yourself.
Facts and explainers
5 things about [TOPIC] that sound made up but are documented.
Category: Top 5 List
Why [EVERYDAY OBJECT] is designed the way it is — and who decided that.
Category: Science Simplified
The [YEAR] decision that quietly changed [INDUSTRY] forever.
Category: History Facts
[NUMBER] psychology tricks interviewers use without telling you.
Category: Psychology Facts
What actually happens in your body in the first hour of [ACTIVITY].
Category: Science Simplified
The three most expensive mistakes people make with [MONEY TOPIC].
Category: Financial Tips
Story-driven
These want the 45-60 second target length and a slower voice. Anything with a twist needs room to land.
The disappearance nobody solved: [CASE], in one minute.
Category: True Crime
He built [COMPANY] out of a garage. Then he lost it in a week.
Category: Business Stories
The night [HISTORICAL EVENT] went wrong — minute by minute.
Category: Wars & Conflicts
A letter arrived forty years late. This is what it said.
Category: Custom Series Brief
The village that vanished between two censuses.
Category: Scary Stories
My grandfather's last voicemail explains everything.
Category: Custom Series Brief
Character and brainrot
The presets at the head of the category grid live here. Keep these at 15-30 seconds — the format is the joke, not the information.
[CHARACTER] vs [CHARACTER]: who actually wins, and why.
Category: Brainrot
The lore of [MEME CHARACTER], explained like a documentary.
Category: Italian Brainrots
A cat narrates the worst day of its owner's life.
Category: Cat Stories
Ranking [THINGS] as if they were transfer signings.
Category: Messi And Ronaldo
If [ANIMAL] wrote a performance review of humans.
Category: Crazy Animals
Things [PUBLIC FIGURE] did before anyone knew their name.
Category: Celebrity Facts
Useful and evergreen
The ones that keep getting views months later. Saves and shares matter more here than watch time.
Five things to do in the first ten minutes of your day.
Category: Productivity Hacks
The [APP OR TOOL] setting nobody turns on — and why you should.
Category: Life Hacks
[CITY] on [BUDGET]: where to sleep, eat and go.
Category: Travel Destinations
How to actually read a [DOCUMENT] before you sign it.
Category: Tutorials
The three specs that matter when you buy a used [CAR].
Category: Car Reviews
What [NEW TECHNOLOGY] will and will not do by next year.
Category: New Technologies
AI Slideshow Questions, Answered
How many images does one video get?
As many as the narration needs. AutoFeed measures the real length of the generated voiceover and renders one image per five seconds, with a minimum of three — so a 45-second story gets nine images and a 20-second one gets four.
Can I write the script myself?
Yes. Pick the Custom Series Brief tile, which is the default and hands you an empty script box, or generate from any preset and then rewrite the script box. The narration, the images and the captions are all built from whatever text is in that box at the moment you render.
Which languages can it narrate in?
There are 39 language and locale options, including English (US, UK, Australia, Canada), Spanish (Spain and Mexico), Portuguese (Brazil and Portugal), French (France and Canada), Arabic (Saudi Arabia and UAE), German, Italian, Japanese, Korean, Chinese, Hindi, Tamil, Indonesian, Turkish, Polish, Ukrainian and Vietnamese.
What does one AI slideshow cost in credits?
It depends on the length you pick, because the length is what drives the spend: the image count comes from the finished voiceover, so a 60-90 second video generates roughly three times the images of a 15-30 second one. With static images a slideshow is 1,000 credits at 15-30 seconds, 1,200 at 45-60 and 1,500 at 60-90. Turning on Animated Images prices the whole video per clip instead — 2,400, 4,800 and 7,200 credits at those same three lengths. The 15-30 second default is the only combination a 1,000-credit free trial can afford.
What is the difference between Camera Shake and Animated Images?
Camera Shake keeps the stills and adds drift so the frame is never completely static. Animated Images replaces each still with a generated five-second clip, and turns Camera Shake off since the clip is already moving. It also changes how the video is priced: instead of 1,000 to 1,500 credits by length, you pay per clip — 2,400 credits at 15-30 seconds, 4,800 at 45-60 and 7,200 at 60-90 — and it takes considerably longer to render.
Which caption style should I use?
YouTuber is the default and the safest across a mixed audience. Match the style to the category rather than to your own taste: Creepy or Glitch for horror, TikToker or Instagramer for lifestyle, and Classic 1 or 2 when the images are busy and the captions need to stay out of the way.
What are the particle overlays for?
They are a full-frame layer — smoke, ember, dust, sakura, glowy, travel or warp — sitting on top of the images. Used sparingly they hide the seam between two generated stills. Used on every video they become the thing people notice instead of the story.
Can a slideshow series run without me?
Yes. Attach the format to a series, pick the days and the whole-hour slots you want, and AutoFeed generates a fresh video and publishes it to your connected TikTok and YouTube Shorts accounts in your own timezone. Plans cap a series at 3, 7 or 14 posts per week.
The Other Generators
Most channels run two or three of these side by side — one that carries the storytelling and one that farms comments.
Reddit Story Video Generator
Real threads, AI narration, word-timed captions and gameplay underneath.
Would You Rather Video Generator
Split-screen dilemmas with a narrated question, a timer and a percentage reveal.
Pick One Video Generator
A POV hook, then numbered photoreal options cut to the beat. No narration.
