Try Mischief 1 free: 10 credits with Google. No card.

Try free

AI Video Generator: Turn Your Images into Short Videos

Last updated: 2026-11-07

An AI video generator search covers several different products. Some invent a clip from a sentence. Some animate a picture you already have. Some build a talking presenter or edit a timeline. This page is the decision page for the tool on this site. It accepts one still image plus a motion prompt and returns a short MP4. It does not create video from text alone. Use the form below if that input matches what you have. If it does not, the sections further down say so before you spend time on an account.

Start from an image you already have

Upload a first-frame image, write a motion prompt and generate here. Sign-in required; failed generations are refunded automatically.

5 credits per clip · 10 free sign-up credits = 2 clips like this. Tip: test at 480p, finish at 720p.

Output: 16:9 widescreen · 3–5 s · 16 fps · MP4 · no audio

What kinds of AI video generators exist

The words "video generator" and "video generator AI" do not name one product. Group the tools by the input they actually require. That split matters more than the brand on the button.

Kind of toolWhat you start withOn this site
Text-to-videoA written scene. No picture required.Not offered. A missing image is rejected.
Image-to-videoOne still, plus words that describe motion.This is the only mode the form runs.
Avatar, lip-sync, or editorA script, a voice, or several shots on a timeline.Not offered. No speech and no multi-shot edit.

Text-to-video

You describe a scene and the model invents every frame: the subject, the place, and the motion. That is useful when you have an idea and no artwork. It is also the wrong expectation to bring here. This generator will not fill in a picture you did not upload.

Image-to-video

You supply the first frame. The picture holds the subject, the colors, and the composition. The prompt only steers what moves and what the camera does. A product photo, a landscape, or an illustration you already like is the right raw material. The result is a short motion study of that frame, not a new storyboard.

Avatar and editor tools

A third group starts from a script or a voice track and produces a presenter, or it asks you to cut several shots together. Those jobs need speech, timing, and usually a vertical frame for short-form apps. This form does not speak, does not burn in captions, and does not assemble a sequence. If that is the job, pick a different category instead of forcing a motion prompt to do it.

What this one does

The form is image-to-video only. You upload one still and write a motion prompt. You then choose a resolution and a length. The request is sent with those fields: the image, the prompt, 480p or 720p, and 3 or 5 seconds. There is no separate control for aspect ratio, frame rate, or an audio track.

You setAccepted values
ImageOne JPG, PNG, or WebP file, 4 MB or smaller
Motion promptRequired text, up to 2000 characters
Resolution480p or 720p
Length3 or 5 seconds
AccountSign-in required before a clip is generated

The published clip is a silent widescreen MP4: 16:9, 16 fps, 832×480 or 1280×720. Those output facts are not extra switches in the form.

A ceramic cup on a wooden table in window light, an example of a single still image to upload
The input is one framed still. The prompt should name the motion (steam, a slow push-in) and what must not move (the cup, the table), rather than inventing a different scene.

Credits scale with the model and settings you pick. On Mischief 1, 480p costs one credit per second and 720p costs two credits per second; MiniMax H3 Max Turbo and Kling 2.5 Turbo cost more per clip. A 3-second Mischief 1 draft at 480p spends fewer credits than a 5-second 720p export of the same prompt. The button shows the cost before you run it. If the generation fails, those credits are returned. A clip you simply dislike is not a failure, and it still spends the credits.

For a closer pass over framing and first-frame choices, walk through a still-image workflow. The same form is also on the home page if you want the rest of that page around it.

What it won't do

  • It will not generate a video from a prompt alone. No image means no clip.
  • It will not add dialogue, music, sound effects, or lip-sync. The file is silent.
  • It will not export a vertical 9:16 or square file. Plan to crop in an editor if a phone feed needs that shape.
  • It will not make a clip longer than 5 seconds, and it will not cut between scenes. One still, one short move.
  • It will not render text overlays you can rely on. If words in the picture must stay readable, say so in the prompt and check the result. Do not expect typesetting.
  • It will not run without an account, and it is not an open-ended free tool. See the trial section below before you assume a price of zero.
You also confirm that you own the image or have permission to use it, and that it does not show a real person without their consent. The generator blocks a submit until that box is checked.

How to make a video with AI in 3 steps

  1. Choose the still. A sharp, landscape frame with one clear subject works better than a busy collage. If the photo is vertical, crop it before upload so the subject stays in the widescreen frame. Keep the file under 4 MB, as JPG, PNG, or WebP.
  2. Write the motion, then stop. Name one camera move and one thing that moves. Name what stays put. A usable line looks like: "Slow drift to the right. Leaves in the foreground shift once. The building stays sharp and does not morph." Wording patterns are collected under motion-prompt examples.
  3. Pick length and resolution, then download. Use 3 seconds at 480p while you test. When the move looks right, run the same image and prompt at 5 seconds or 720p. Download the MP4. Add audio afterward if the destination needs sound.

Sign in before the first run. The cost on the button is the number of credits that attempt will use. Leave the page if you want; a finished clip stays on the account. A hard failure returns the credits. You do not file a ticket for that ordinary refund.

Is it a free AI video generator online?

It is online: nothing to install, and the form runs in the browser. You still need an account, and it is not an open-ended free generator. A new Google sign-up can include a small free trial of credits. An email-and-password sign-up does not start with that trial. The trial is there so you can judge a short clip. It will not carry a batch of exports.

Do not treat this page as a quote of the current grant. The amount, the eligibility, and what happens when it is used up are maintained on the current free-trial conditions and on current prices and plans. After the trial, you buy a credit pack or a plan. Because 720p costs twice as many credits per second as 480p, draft on the cheaper setting first so a trial goes further.

Searches for a free AI video generator online often expect no account and no cap. This tool does not match that expectation. It matches a narrower one: try a limited number of image-and-motion clips, then pay for more seconds.

Who it's for, and who should skip it

Use it if

  • You already have a product photo, an illustration, or a landscape and you want a few seconds of motion from that exact frame.
  • You are comparing AI video generators and need one that starts from an image rather than from a blank prompt.
  • A silent widescreen MP4 is acceptable, and you can add music in an editor or post the clip without sound.

Skip it if

  • You only have a sentence and no picture.
  • You need a talking presenter, burned-in captions, or a native vertical file.
  • You need a minute-long edit with several shots, cuts, or a voice track.
  • You need generation with no account.

If you are in the first list, the form at the top of this page is the tool. If you are in the second list, another category of video generator will waste less of your time.

ClipWhim is an independent product and is not affiliated with, endorsed by or sponsored by The Biological Computing Co. (TBC). “Mischief 1” is used only to describe the third-party model ClipWhim runs, accessed via API; the name belongs to its owner.

Frequently asked questions

Can this AI video generator make a video from text only?
No. The form requires one image file plus a motion prompt. A text-only submission is rejected before a clip is made. If you do not have a picture, this is the wrong kind of tool.
What should the motion prompt describe?
Camera and motion, not a new scene. The still already shows the subject, the lighting, and the framing. Say what moves, how fast, and what must stay still. One action is enough for a 3- or 5-second clip.
Is it free to use online?
You can open it in the browser, and you do need an account. A new Google sign-up can include a small free trial of credits. An email-and-password sign-up does not start with that trial. It is not an open-ended free tier. Current conditions are on the free-trial page and the pricing page, and paid credits are required after the trial is used up.
What file do I download?
A short MP4. The form only offers 480p or 720p and 3 or 5 seconds. The published output is a silent widescreen clip, so you add music or voice later in an editor if you need sound.
Will a vertical phone photo work?
You can upload a JPG, PNG, or WebP up to 4 MB. The clip is widescreen, and this form has no vertical export. Crop the photo yourself first if the top and bottom edges matter.
How is this different from the other guides on this site?
This page is for choosing a kind of AI video generator: what this tool is, what it refuses to do, and who it fits. The linked guides cover the still-image workflow, motion-prompt wording, the current free trial, and prices. They do not change the input this form accepts.