Pixwhim

Text to Video AI

Type a scene and get a short AI video. No camera, no footage, no editing software.

  • from 45 credits
  • Free account, 20 credits to start
  • Uses purchased credits
1 · Model
3 · Options

Video uses purchased credits. Free credits work for image tools.

Example
“Close-up of a small paper boat sailing down a rain-soaked city street gutter, reflections of neon signs in the water, cinematic slow motion”

See what it does

Paper boat in the rainPrompt: “Close-up of a small paper boat sailing down a rain-soaked city street gutter, reflections of neon signs in the water, cinematic slow motion”

How it works

  1. 1

    Describe your scene

    Write what should happen, where, and how the camera moves, like 'a red fox walks through fresh snow at sunrise, slow tracking shot'.

  2. 2

    Pick model, length and quality

    Use MiniMax H3 Max Turbo for 480p or 768p clips of 5 or 10 seconds, or Veo 3.1 Lite for 720p clips of 4, 6 or 8 seconds.

  3. 3

    Generate and download

    The AI renders your clip. Download it, or tweak the prompt and try another take.

Why people use it

Video from words alone

No photo or footage needed. Describe a scene and the AI creates it from scratch.

Two video models

MiniMax H3 Max Turbo is the default. Google's Veo 3.1 Lite makes 720p clips. Both generate sound.

Flexible length and quality

MiniMax H3: 480p or 768p, 5 or 10 seconds. Veo 3.1 Lite: 720p, 4, 6 or 8 seconds.

Direct the camera

Ask for zooms, pans, tracking shots and close-ups in plain English.

Clear pricing

Videos cost from 45 credits, depending on the model, length and quality. Paid with purchased credits.

What text to video AI is good for

Text to video AI turns a written description into a short video clip. It's handy when you need footage that would be hard or expensive to shoot: an establishing shot for a story, a background loop for a presentation, a mood clip for a social post, or a quick visual to test an ad idea before a real shoot.

Two models are available. MiniMax H3 Max Turbo is the default, with 480p or 768p output in 5- or 10-second clips. Google's Veo 3.1 Lite creates 720p clips of 4, 6 or 8 seconds. Both generate sound, which is useful when you want ambience or effects without adding audio yourself.

How to write a good video prompt

A clear prompt usually covers four things:

  • Subject: who or what is in the shot ("an elderly fisherman", "a glass perfume bottle").
  • Action: what happens ("mends a net", "rotates slowly on a pedestal").
  • Setting and light: where and when ("on a wooden dock at dawn, soft fog").
  • Camera: how it's filmed ("close-up, shallow depth of field, slow push-in").

Keep each clip to one simple idea. Short videos can't fit a whole story, so plan several clips and edit them together if you need more.

Text to video or image to video?

Text to video invents everything from your prompt. If you need a specific look, such as your own product, your face or a consistent character, start with an image instead. Generate one with the AI Image Generator or upload a photo, then use Image to Video so the clip starts from that picture.

Credits and limitations

Video uses purchased credits only, at from 45 credits depending on your choices. A short 480p test is a cheap way to try a prompt before making a longer version. Like all AI video today, results can include warped hands, flickering details or unreadable text. If a take doesn't work, simplify the action or rephrase the prompt.

Frequently asked questions

Try next