How to make AI videos: from a photo or an idea to a finished clip with sound
To make an AI video, start from a photo or a written idea, choose how long the clip runs and what shape it is, and generate it: in PrismPoster the clip comes back with sound. Then add a song, extend the clip from its last frame, and export a finished Short from the browser editor. This guide walks through each step, with the real limits at every one.
The short answer
Upload a photo or describe a shot, pick a length of 4, 6, 8, 10 seconds and a 16:9 or 9:16 frame, and press Generate. The clip renders with sound. For longer videos, extend a clip from its last frame or cut several clips together in the editor, add a song in any of 35 languages, and export.
Step 1: choose where the video starts
There are two starting points, and they suit different jobs.
A photo. Upload a JPEG, PNG or WebP on the photo to video page and describe the motion you want: "slow push-in, her hair lifts in the wind", "the dog runs toward the camera through the surf". The picture becomes the first frame, so the subject, the light and the composition stay yours. This is the fastest way to a good result, because the model does not have to invent what the scene looks like.
A photo of another real person renders only after you confirm you have their permission, and the clip carries a visible AI label. Photos of children and of public figures are refused when you attach them. Your own face, once you verify it, can appear in clips from the Premium engine.
An idea. Write the shot in a sentence or two on the AI video generator with sound: who or what is in frame, what moves, where the camera goes, and the light. Good prompts name the camera move and one clear action rather than a list of adjectives. For a longer scene, write it beat by beat: what happens in the first seconds, then what changes.
Step 2: pick the length, the frame and the resolution
The Standard engine renders clips of 4, 6, 8, 10 seconds, in 16:9 for YouTube and the web or 9:16 for Shorts, Reels and TikTok. An account that has never paid renders at 360p or 720p; 1080p and 4K open after any plan or credit pack.
The Premium engine, included on every paid plan, renders one continuous take of up to 30 seconds in 5 frame shapes, up to 4K, with the sound rendered together with the picture. It suits a single moving shot that should not cut; the Standard engine suits short clips you plan to edit together.
Step 3: generate, and know what it costs
Generation is paid in credits from one wallet. The shortest Standard clip costs 30 credits at 360p and 80 credits at 720p, and a longer clip costs more per second. The pricing page lists every rate.
New PrismPoster accounts get 200 free starter credits in many countries, with no card. The grant is not offered in Belarus, where the first credit pack is the entry, and unused starter credits expire 24 hours after they are granted.
A render that fails costs nothing: its credits return to the wallet.
Step 4: add sound and a song
Standard clips come back with sound: ambience, movement and effects that match the picture. To score a video with music, make a song on the AI song generator, with your own lyrics or a draft from the studio, in any of 35 languages. Every song comes with vocals, separated stems and cover art, and you can use it commercially on every plan.
For a music video built around a song you already have, the lyric video maker turns the track and its words into a vertical lyric video, and the stem splitter separates the vocal from the instrumental.
Step 5: extend, cut and export
Open the clip in the browser timeline editor. From there you can:
- Extend it. The editor continues a Standard clip from its last frame, up to 40 seconds in total, so a moment can run on instead of stopping.
- Cut several clips together. Drop clips on the timeline, trim them and put them in order; the editor splits each clip's audio onto its own lane so you can mix it under your song.
- Export. Choose 16:9, 9:16, 1:1 or 4:5 and export up to 4K, then download the file and post it yourself.
Labels and rights
PrismPoster writes machine-readable metadata marking every file it generates as AI-generated, on a best-effort basis, so the platforms you post to can label it. Images and videos made on a free account also carry a small visible AI label. You can use what you make commercially on every plan, including the free one.
Frequently Asked Questions
How do I make an AI video for free?
In the countries where starter credits are offered, a new account gets 200 free starter credits with no card, enough for short clips at 360p or 720p; unused ones expire 24 hours after they are granted. Where the grant is not offered, a credit pack is the entry.
How long can an AI video be?
A Standard clip runs 4 to 10 seconds and extends from its last frame up to 40 seconds. A Premium take runs up to 30 seconds without a cut. Anything longer is several clips cut together in the editor.
Can I make an AI video from my own photo?
Yes. Upload a JPEG, PNG or WebP and describe the motion. A photo of another person needs their permission and carries an AI label; photos of children or public figures are refused.
Do AI videos have sound?
Yes. Standard clips come back with sound that matches the picture, and Premium takes render the sound together with it. You can add a song on top in the editor.
Turn ideas into high-definition AI video clips & timeline cuts
Generate text-to-video and image-to-video in 360p up to 4K, extend scenes per second, and assemble complete multi-track videos with auto-captions and zero desktop install.