Features
AI Video Generator Features
Everything AI Shorts Studio does between your script and the finished MP4, with the exact limits, styles, voices and export settings.
Script analysis and scene planning
Your script is the starting point for everything. The generator splits it into sentences, checks it against the word budget for the length you chose, and divides it into scenes at natural sentence breaks so no scene starts or ends mid-thought.
Narration is paced at up to about three words per second, with a short pause before the voice starts and a moment of silence at the end. That gives each video length a fixed word budget:
| Video length | Maximum words | Scenes |
|---|---|---|
| 5 seconds | 12 | 3 |
| 10 seconds | 25 | 3 |
| 15 seconds | 40 | 4 |
| 20 seconds | 55 | 4 |
| 25 seconds | 70 | 5 |
| 30 seconds | 85 | 6 |
The word count and estimated narration time update as you type. If your script is over the budget, nothing is cut silently: you get a suggested shorter version that keeps your opening sentence (the hook) and your last sentence (the call to action), and you decide whether to use it.
Scene visuals and motion
Every scene gets its own full-frame visual, created by the app's built-in visual engine. The engine reads the keywords in that scene's text, such as "family", "city", "code" or "sunrise", and combines them with the colour palette and mood of your visual style. No stock footage is used.
During rendering each visual is animated with slow camera moves (zoom in, zoom out or a gentle pan) and joined to the next scene with the style's transition: a crossfade, a fade through black, a slide or a zoom.
10 visual styles
A style sets the look, camera motion, transitions and narration pacing in one choice, and pre-selects a matching caption style and music mood. You can still change the captions and music before generating.
| Style | Look and feel | Default captions | Default music | Transition |
|---|---|---|---|---|
| Emotional Documentary | Slow, warm and intimate with gentle zooms and soft piano. | Bold White + Yellow Highlight | Soft Emotional Piano | Crossfade |
| Social Awareness | Serious, high-contrast visuals with steady pacing. | Bold White + Yellow Highlight | Cinematic Ambient Score | Fade through black |
| Motivational Story | Rising energy, sunrise tones and uplifting music. | Impact Headline | Uplifting Inspiration | Zoom |
| Technology Explainer | Clean, cool tones with crisp slides and synth pulse. | Boxed Word Highlight | Tech Pulse Synth | Slide |
| AWS Tutorial | Cloud-themed visuals, orange accents, clear captions. | Boxed Word Highlight | Tech Pulse Synth | Slide |
| Angular Tutorial | Developer aesthetic with red accents and code motifs. | Boxed Word Highlight | Tech Pulse Synth | Slide |
| Business Advertisement | Polished, bright and confident with punchy cuts. | Boxed Word Highlight | Bright Corporate | Zoom |
| Cinematic Storytelling | Teal-orange film look, slow drift and deep score. | Impact Headline | Cinematic Ambient Score | Crossfade |
| Educational Short | Friendly, bright and easy to follow. | Bold White + Yellow Highlight | Uplifting Inspiration | Crossfade |
| Product Promotion | Energetic, glossy product framing and fast pacing. | Impact Headline | Bright Corporate | Zoom |
AI voiceovers
Narration is generated with Piper, a neural text-to-speech engine that runs on the AI Shorts Studio server. Pacing follows your style: documentary and cinematic styles speak slightly slower with longer pauses between sentences, while promotional styles are a little quicker.
| Voice | Accent | Licence |
|---|---|---|
| Ryan | US English, male (high quality) | CC BY-NC-SA 4.0 dataset: non-commercial use only |
| Alan | UK English, male | Mimic3 voices dataset: see the model card |
| Amy | US English, female | Mimic3 voices dataset: see the model card |
| Lessac | US English, female | Blizzard 2013 dataset licence: check before commercial use |
Each voice is trained on a dataset with its own licence, and some allow non-commercial use only. The licence is shown in the generator when you select a voice. Check it before using a video commercially.
Automatic captions and subtitles
Captions are timed to the narration and appear a few words at a time, with each word highlighted as it is spoken, so viewers can follow along with the sound off. They sit in a mobile-safe area above the bottom of the frame, where Shorts, Reels and TikTok overlay their own buttons and text.
Bold White + Yellow Highlight
Heavy uppercase words, up to 3 on screen, with the spoken word in yellow.
Impact Headline
Large condensed uppercase type, up to 3 words, for high-energy videos.
Boxed Word Highlight
Up to 4 words in sentence case; the spoken word sits in a yellow box.
Clean White (no highlight)
Up to 5 words in plain white with an outline, for a calm, minimal look.
Every video also comes with an SRT subtitle file that you can upload to YouTube or other platforms as closed captions.
Background music and audio mix
Choose one of five music moods or no music at all. The music is composed by the app for each video, not taken from a stock library.
Soft Emotional Piano
Gentle and reflective.
Uplifting Inspiration
Bright and encouraging.
Cinematic Ambient Score
Wide, atmospheric pads.
Tech Pulse Synth
Steady electronic pulse.
Bright Corporate
Clean and positive.
No music
Voice only.
When the narrator speaks, the music is automatically turned down (ducked) and comes back up in the pauses. The final mix is loudness-normalised to about −14 LUFS, a common reference level for online video, so your video is not noticeably quieter or louder than others in the feed.
Formats and export
| Format | Aspect ratio | Resolution | Best for |
|---|---|---|---|
| Vertical | 9:16 | 1080×1920 | YouTube Shorts, Instagram Reels, TikTok |
| Landscape | 16:9 | 1920×1080 | Regular YouTube videos, websites, presentations |
| Square | 1:1 | 1080×1080 | Instagram and Facebook feeds, LinkedIn |
Videos are exported as MP4 files with H.264 video at 30 frames per second and AAC audio, which every major platform accepts. There is no watermark. Download the MP4 and the SRT captions as separate files.
Live progress, preview and project library
After you select Generate Video, each stage is shown as it happens: script analysis, scene generation, scene visuals, voiceover, captions and music, and final rendering. The server renders one video at a time, in order, and rendering carries on if you close or reload the page.
Finished videos play in your browser. From there you can download them, regenerate with the same settings, or start a new video. Every video appears in the project library in the generator, where you can reopen or delete it.