Katama
動 AI video generator

Prompt in.
Film out.

Every top video model behind one box — with sound, in any ratio. Type the shot, pick an effect, and Katama routes it to the model that can actually deliver it.

rain-slick Tokyo alley at night, neon reflections, slow dolly in
Generate
See what came out
Start free — 3 test videos, no card. Hover any clip for sound →
Rendered live in your browser with Remotion — every frame is React, not a video file.
Seedance 2.0Kling 3.0 ProVeo 3.1Hailuo 2.3+ more

Made on Katama. Hover any for sound.

Browse the gallery →
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound

Effects that do the directing for you.

Pick one and it leads the shot — camera move, timing and look go into the prompt before your words do. These are the 12 that ship in Cinema today.

A still you already have.
Moving, in one pass.

Drop a product shot, a portrait, a frame you liked. Image-to-video runs the same model families image-conditioned — your subject stays your subject, and the price per second is the same as text-to-video.

  • Your exact product, not a lookalike
  • Same credits per second as text-to-video
  • Any ratio out — 9:16 for Reels, 16:9 for YouTube
Still input
Still in
🔊 sound
Clip out

Product videos, without the shoot day.

One photo of the thing you sell is the whole production budget. Pick the scene, let it run in batch, and keep the takes that convert — for the shop page, the ad account and the feed, in the ratio each one wants.

01Upload the product shotThe one from the shop page is enough.
02Choose the scene and the moveStudio, table-top, lifestyle, hand-held — plus an effect if the shot needs a hook.
03Batch, then keep the winnersSpin variants of the same product and cut the ones that don't land.
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
🔊 sound
Built forE-commerceBeautyFood & drinkFashionSupplementsAppsOpen the marketing studio →

One box. The whole catalog.

No six subscriptions, no ten tabs. Katama routes your shot to the model that fits it — and shows the credit cost before you click, from the same source that charges you.

Seedance 2.0
ByteDance · up to 4K
Kling 3.0 Pro
3–15s native
Veo 3.1
Google · premium tier
Hailuo 2.3
MiniMax · fast
Wan 2.6
Alibaba
LTX-2.3
Lightricks · cheap tier
Ratios21:916:94:31:13:49:16· 4–15s per clip · with sound
01
Describe the shot
A line of text, or drop the still you want animated.
02
Katama routes it
Model, settings and effect picked for that shot — not a lucky roll.
03
Get a finished clip
With sound, in your ratio, ready to post or drop into Cinema.

Video questions, answered.

Which models can I use?+

Seedance 2.0, Kling 3.0 and 3.0 Pro, Veo 3.1, Hailuo 2.3, Wan 2.6 and LTX-2.3 — plus the rest of the catalog, behind one box. No separate subscriptions.

What does smart routing do?+

You describe the shot; Katama picks the model and settings that fit it. You stop burning credits finding out which model was the right one.

Can I animate an image I already have?+

Yes. Image-to-video runs the same model families image-conditioned, at the same price per second as text-to-video — your subject stays your subject.

Do the clips have sound?+

Yes — voice, music and effects are built in, so what comes out is a finished clip, not a silent loop.

How long can a clip be?+

4 to 15 seconds per generation depending on the model. Chain them in Cinema when you need a whole film.

Is it free to start?+

3 test videos free, no card. Premium models run on credits, and the credit cost is shown before you click from the same source that charges you.

Stop describing it. Watch it.

Three test videos free, no card. Then scale onto the premium models when the shot has to land.

Generate a video →