Text to video
Write the shot in plain language. The prompt is composed in briefing order — subject, framing, camera, optics, light, grade — so the model receives direction the way a crew would, rather than a pile of adjectives.
Back to home
AI video generator
Describe a shot and get video, or animate a still you already have. Pixel Engine runs 16 video models behind one interface, and adds the part most generators leave out: direction. You choose the camera move, the lens, the depth of field, the lighting and the film stock, instead of hoping the model guesses.
Image to video
Start from a still and animate it. 14 models accept a start frame, and some accept an end frame too, so a shot can be made to land on a specific image.
Direction, not just prompting
Every control below is a real option in the studio, applied to the prompt before it reaches the provider.
42 camera moves
Dolly, crane, orbit, handheld, whip pan, bullet time and more. Up to three can be stacked in one shot.
12 lenses
Anamorphic, ultra wide, portrait, telephoto, macro, tilt-shift — plus aperture for depth of field.
11 film looks
ARRI Alexa, RED, Sony Venice, IMAX 70mm, Kodak 16mm, Super 8, VHS and others.
Video models and what they cost
| Model | Credits | Durations | Notes |
|---|---|---|---|
| LTX 1 | 5 | 5s | text-to-video |
| Kling 1.6 | 8 | 5s / 10s | image-to-video |
| Kling 1.6 Pro | 13 | 5s / 10s | image-to-video |
| Hailuo 02 | 8 | 6s / 10s | image-to-video |
| Seedance Pro | 17 | 5s / 10s | image-to-video |
| Veo 3.1 | 22 | 4s / 6s / 8s | image-to-video · start + end frame · native audio |
| Veo 3.1 Fast | 12 | 4s / 6s / 8s | image-to-video · native audio |
| Wan | 11 | 5s | image-to-video |
| Kling 3 Pro | 16 | 5s / 10s | image-to-video · native audio |
| Kling 3 | 12 | 5s / 10s | image-to-video · native audio |
| Kling 2.6 Pro | 11 | 5s / 10s | image-to-video · native audio |
| LTX 2 | 10 | 6s / 8s / 10s | image-to-video |
| LTX 2 Fast | 9 | 6s / 8s / 10s | image-to-video |
| Kandinsky 5 | 6 | 5s / 10s | text-to-video |
| Veo 3.1 Lite | 6 | 4s / 6s / 8s | image-to-video · start + end frame · native audio |
| Wan 2.6 | 14 | 5s / 10s | image-to-video |
6 of these generate sound natively, so the clip arrives with audio rather than needing a separate pass.
After the render
Generated clips land in your library, and the browser-based clip editor trims them, cuts silence, adds music under your voice and burns in captions — without uploading your footage anywhere.
Make the next render easier to direct
These practical playbooks turn an idea into clear, editable shots before you spend credits on a full sequence.