50% OFF Your first purchase is half price — the discount is already in the price you see. Ends October 30.
Text or a photo in, a finished AI video with sound out, from the models the pros reach for. No app, no editor.
The newest Seedance tier. Text or photo to video at 480p, 720p or 1080p, priced per second.
Generate video from text or images. 480p–1080p, with audio.
Cinematic video. Lite and Fast tiers, 720p–4K.
Text or photo to video at 720p, 1080p or 4K, audio optional. Priced per second.
Kling's next model: 30-second clips, up to 4K, 15 references. Flash in early access, full release in October.
Mid-weight text or image to video. 480p–1080p, audio optional, priced per second.
Text or photo to video at a fixed 2K, 5 to 10 seconds. Priced per second, no tier to pick.
Black Forest Labs' video model. Text or photo to video with sound, up to 20 seconds at 720p or 1080p.
The cheapest Seedance 2 tier. 480p or 720p video, priced per second.
The middle Seedance 2 tier. Video from text or a still image at 480p or 720p, priced per second.
Text or photo to video, 5 or 10 seconds, with generated sound optional.
The fast Kling 3.0 tier. Text or photo to video at 720p or 1080p, priced per second.
Open-weights video with synchronised audio — up to 4K, 20 seconds, 50fps.
Alibaba's Wan 3.0 — up to 30 seconds in one take at 480p, 720p or 1080p, with audio in the same pass.
One game photo in: the game face, then a hard cut to the same player driving a kids’ toy car.
One selfie in: you film yourself at a school blockade, the riot police charge, and you run.
One photo in: you pop the bottle, the spray shoots up and the frame freezes with Happy Birthday on it.
One photo in: you lean back, point at the lens and nod on every beat. Add more photos for a lineup.
One photo of you two in: the run-up, the catch and the Dirty Dancing lift, in your own room.
One group photo in: your whole crew steps at the camera in sync, then stomps right over the lens.
One photo in: you doing the zombie dance on a foggy street full of jack-o'-lanterns.
One photo in: you mouthing the hook, pointing at the lens and flipping your hair like a film star.
One photo in: you in the living-room crew, squatting low and beating your chest on the drop.
One photo in: you doing the Dame Hombro freestyle, shoulders bouncing on the dembow beat.
One photo in: you doing the Wipe Me Down dance, dropping low and flicking your hands at the floor.
One photo of your car in: black liquid grabs the badge and builds your car back, piece by piece.
Write the subject, motion, camera and mood in one prompt, and the AI video generator turns it into a moving shot.
Start from a still, or pin the first and last frames on Seedance 2.0 and Veo 3.1, and the AI video lands where you pointed it.
Seedance 2.0, Veo 3.1, Wan 3.0 and LTX 2.5 write the sound in the same AI video pass, so the clip comes back with audio already in it.
Veo 3.1, Kling 3.0 and LTX 2.5 reach 4K, and Wan 3.0 runs a full 30-second AI video in one take.
Fourteen AI video models on one balance: Seedance 1.5 Pro or Seedance 2 Mini for cheap drafts, Veo 3.1 or Kling 3.0 for the final cut.
Put subject, action, camera, light and style into one clear scene; the AI video generator follows what you write.
Choose 480p to 4K and the clip length; the credit estimate updates as you change either.
Confirm the estimate, generate, and download your AI video.
Fourteen AI video models run here: Seedance 2.5, Seedance 2.0, Seedance 2 Fast, Seedance 2 Mini and Seedance 1.5 Pro, Veo 3.1, Kling 3.0, Kling 3.0 Turbo and Kling 2.6, MiniMax H3 and MiniMax H3 Fast, FLUX 3, LTX 2.5 and Wan 3.0.
Yes, with limits. The Free engine tab on this page makes video with no account at all. Sign in and you get 20 free credits plus daily check-ins worth up to 100 credits a week; a free account runs MiniMax H3 Fast at 480p, where a 4-second clip is 10 credits. Any purchase, a credit pack or a plan, opens 768p on that model and every other model on this page.
It depends on the clip you need. For the longest shots, Seedance 2.5 and Wan 3.0 go to 30 seconds. For 4K, Veo 3.1, Kling 3.0 and LTX 2.5. For sound generated with the picture, Seedance 2.0, Veo 3.1, FLUX 3, LTX 2.5 and Wan 3.0. For free, MiniMax H3 Fast. All of them run from this one page on one credit balance, so you can try the same prompt on two models and keep the better take.
Text to video and image to video, with start and end frames on the models that take them. Every AI video model here accepts a written prompt.
Yes. Seedance 2.0, Veo 3.1, FLUX 3, LTX 2.5 and Wan 3.0 generate synced audio with the AI video; on Kling 3.0, Kling 2.6 and Seedance 1.5 Pro sound is an option that changes the price.
Each AI video model bills its own way: most per generated second, Veo 3.1 per video, all by resolution. The credit estimate always shows before you generate.
Four very different AI video pieces: a wildlife documentary shot, an illustration drawing itself, a minimal animation, and a camera move through a harbour city.
Moss-Coated Sloth and Cub on a Coral Reef
An 1890s Field Guide Plate Drawing Itself
A Small Character and an Unspooling Thread
Rain-Soaked Harbour Market and the Great Gate
Any plan or pack — even a $19.90 one — unlocks a clean copy of every image you have made here, starts each new run instantly, and refunds one that fails.
Pay safely and securely with
Payments are processed by Stripe over an encrypted checkout. Your card details are handled by Stripe and are never stored by Kavel.
Your 4-second clip is free every day you sign in. The day one deserves the full take, longer (up to 30 seconds), sharper (up to 4K) and with its own sound, your first plan or pack is half price, and only the first.
Open the generator
