AI video that stays editable.

Every cut, caption and move is a layer you can change, no editing skills needed. Approve one video and it becomes the template for every version.

Start free in Studio Watch real examples

Also works inside

3,000+ companies create with JsonCut
1M+ render seconds every month

From a brief to finished images and videos.In one project.

Build one result hands-on or together with AI. Keep the project editable, then finish one version or connect it to the workflows that need many.

01
Bring the idea
Start with a brief, your media, a product page or an existing design.
02
Create your way
Build it hands-on, or let the AI Director draft it and keep changing it with you.
- In Studio
- With AI
03
Keep one editable project
Design, media, motion and sound stay editable for the next change.
04
Choose what may change
Reuse the approved design with new text, media, products or languages.
05
Publish one or many
Export in Studio, or create variants from your app, a workflow or an AI agent.

POST …/templates/launch/bulk-renders
{ "items": [
  { "values": { "title": "Sale" } },
  { "values": { "title": "Soldes" } }
] }

Programmatic when work repeatsAPI, MCP and templates render every version.

Real projects.Posts ready to publish.

Made in JsonCut, and every caption, title and shot in them is still a layer you can change. The creator reel is the AI Director’s work from a single sentence.

Creator reels

Two raw clips and one sentence. The AI Director Pro wrote the voiceover, timed every caption word and added the timer, the three pillars, stickers, music and sound, in about six minutes.

The requestTurn these two raw clips into a 15-second vertical reel for my creator account about planning a week of content in ten minutes. Write a short punchy voiceover, bold word-by-word captions, motion graphics overlays on the key points, punch-in zooms on the beat, music and sound effects.

Try this prompt yourself: Creator reels

Motion graphics and 3D

A 60-second launch film with a 3D wall of AI outputs, kinetic type, voice, captions and 54 sound effects. Open it in Studio and every layer is yours.

Open the launch film: Motion graphics and 3D

Product promos

A promo for a fictive UGC-ad tool: a brief is typed, real phone clips are scanned and the ad assembles itself, with captions and an offer card. Every clip and line can be swapped.

Open the promo: Product promos

Talking-head explainers

A creator at his desk explains one tip. Glass cards, a retention chart and word-by-word captions land on the words he says, and every card is a layer you can move or retype.

How talking heads work: Talking-head explainers

Everything a video needs.In one studio.

The best AI models for video, images, voice and music, a library of graphics and sound, and an editor that keeps all of it in layers you can change. You ask, the AI Director picks the right tool.

Veo 3.1Google“Slow dolly along classic cars at a drive-in cinema at dusk, the screen glowing warm orange”4 s · 720p · text to video

Kling 4Kuaishou“Matte black headphones turning on a dark reflective surface, an orange light streak sweeps across”5 s · 720p · text to video

AI videoVeo 3.1 · Kling 4 · Seedance 2.5 · Wan 3.0 · Luma Ray 3.2

See all 13 video models

Say what you want.The AI Director does the craft.

Ask in your own words. It plans the scenes, makes what is missing with the right model, cuts on the beat and times every caption. A change is one more sentence, and it lands as a layer you can still touch.

Start with one of these

Video, reels, images, carousels.One editor.

Start from a sentence, a template or a blank canvas. Whatever the AI made sits in the same editor as what you add yourself, so you never export to another tool to finish.

The video editor

One approved ad.Every market, product and format.

Save any design as a template and choose what may change: the product photo, the colour, the headline, the price. Layout, type and motion stay locked, so every version looks like the one the brand signed off.

How bulk versions work

1 template, filled from any source, becomes hundreds of versions in one bulk job, in every format and language. Rows per job by plan

Every layer stays yours.Change one word, not the whole video.

What the AI makes arrives as normal layers: titles, captions, clips, music. Click the headline, type, and the frame changes while you write. Colour, size and motion work the same way.

25+image and video models in one place
21voices for voiceovers, each in 15 languages
480+sound effects, placed on the beat
Every wordtimed in auto captions, ready to restyle

Made in JsonCut.Open any of them and make it yours.

Launch films, UGC ads, explainers, reels. Every example opens as an editable project, with its layers, timing and media.

Browse all examples

Build for free.Pay when you publish.

Building and previewing in Studio costs nothing and needs no card. To export real work and use the AI Director, start the 14-day trial, then pick a plan.

14-day trialEverything in Pro: exports, AI media and the AI Director. Card required; cancel before day 14 and pay nothing.$0

HobbyPublish personal work and run smaller automations.from$10/monthbilled yearly

ProProfessional creation, AI media and recurring production.from$29/monthbilled yearly

BusinessContinuous production for teams and high-volume workflows.from$49/monthbilled yearly

Compare plans and credits

Before you start.The questions we hear most.

Do I need editing or prompting experience?

No. Describe what you want the way you would tell a colleague, or just drop in your footage. The AI Director plans the video, cuts it, adds captions, graphics, music and sound, and checks every frame before it hands the result back.

Can I change what the AI made?

Yes. Every caption, title, clip, graphic and sound arrives as a normal layer on the Studio timeline. Change one word, swap the music or retime a scene without starting over, by hand or by asking the Director.

Can I use my own footage?

Yes. Upload video, photos and audio. The Director can cut a speaker’s pauses and filler words, reframe landscape footage for 9:16, steady shaky clips and add captions timed to every word.

How do templates keep every version consistent?

You choose which parts may change, such as the headline, photo, colour or language. Layout, timing and motion stay exactly as approved. Fill the fields from a table, a form, the API or an assistant like ChatGPT or Claude, and preview each version before it renders.

Which AI models can I use?

Image, video, music and voice models from Google, OpenAI, Kling, ByteDance, Black Forest Labs, ElevenLabs and more, in one studio. Voiceovers come in 21 voices and 15 languages. Every model shows its price in credits before you generate.

What does it cost to try?

Building and previewing is free. Exports, AI generation and the AI Director are included in the 14-day trial, so you can publish real work before you choose a plan.