KinoViz AI Assistant

Overview

KinoViz AI is a conversational design assistant built into the editor. Describe what you want in plain language and it builds animated scenes, edits the elements already on your canvas, adds motion, generates image assets, and more. Everything it produces stays on your canvas and remains fully editable by hand afterward.

The KinoViz AI Assistant is currently paused. The rest of this page describes how it works and is kept for reference; no plan includes AI generation right now.

Two Ways In

From the dashboard

On the studio dashboard there's a Describe your animation box at the top. Type a description (or pick one of the suggestion chips like Title reveal, Chat bubbles, or Brand showcase) and press send. KinoViz creates a new project, opens the editor, and hands your prompt straight to the assistant — you land in the editor watching your animation take shape.

From the editor

Inside any project, open the KinoViz AI panel from the right side of the editor:

1. Click the AI button (sparkle icon) in the top bar, or press Cmd/Ctrl + J.
2. Type your request in the message box and press send.
3. Watch the assistant work — it narrates each step and updates the canvas live.

Press Cmd/Ctrl + J again (or the close button) to hide the panel without losing the conversation.

What It Can Do

KinoViz AI works across the whole editor, not just first-shot generation:

Because every result is made of normal KinoViz elements and keyframes, you can select anything the assistant created and edit it exactly as if you'd drawn it yourself.

A Conversation, Not a One-Shot

KinoViz AI is designed as a back-and-forth. For anything substantial — a landing hero, a pricing scene, a multi-element layout — it works plan-first:

1. It proposes a plan. The assistant reads your request, picks a visual direction, and shows a short plan card describing what it intends to build.
2. You decide. Click Build it to proceed, Tweak this to type changes to the plan, or Cancel to drop it.
3. It builds. Once you approve, it assembles the scene step by step.

Small edits ("make the title red", "snappier entrance") skip the plan and happen immediately. After anything is built you can keep the conversation going — ask for changes, or switch to editing by hand at any point.

Polish (Pro+)

After the assistant generates something, Pro and Studio users can click Polish — the assistant reviews the canvas visually and auto-fixes issues like weak contrast or misaligned elements.

Undo and New conversation

• Undo rolls back the last AI change; right-click it to undo all changes from the session.
• New conversation clears the chat and starts fresh (your canvas is untouched).

Attachments

You can attach up to 3 files to a message:

Attach an image as a style or layout reference (the "recreate this as an animation" flow), or attach a document to give the assistant a brief, script, or copy to work from. Drag files onto the panel or use the attachment button next to the send box.

Credits and Billing

KinoViz AI spends credits from your balance, and the exact cost is shown before you send — a small chip next to the send button reads Free, 1 credit, or N credits, with a tooltip breaking down where the cost comes from.

Default per-message costs (an admin can adjust these — always trust the on-screen preview):

No charge for chat. Greetings, questions, and small talk don't cost anything — you're only billed when the assistant does real work. If a task fails for a system reason, the credits are automatically returned.

While the AI Assistant is paused, no plan includes AI credits and credit packs are not sold. Credits already on your account are kept.

See Billing & Plans for the full credit-pack and plan details.

If You Disconnect

If you refresh the page or lose your connection while a reply is still generating, come back to the same project — KinoViz AI reconnects and resumes the in-progress response so you don't lose the work.

Writing Good Prompts

The clearer your description, the better the result.

Be specific

Good: "A 16:9 social media ad. Blue gradient background. White bold title 'Summer Sale' slides in from the left at 0.3s. Subtitle '50% Off Everything' fades in at 0.8s below the title. A yellow star bounces in from the right at 1.2s."

Vague: "Make a cool ad animation."

Name your elements

Mention the kinds of things you want — title, subtitle, shape, rectangle, circle, star, speech bubble, image, 3D model. The assistant maps these to the right element types.

Describe timing

• Timestamps: "at 0.5 seconds", "after 1s", "at the end"
• Sequence words: "first… then… finally"
• Together: "simultaneously", "at the same time"

Specify style

• Canvas: "16:9", "square", "portrait"
• Color: "blue background", "white text", "gradient from #007aff to #00c6ff"
• Type: "bold", "large", "italic"
• Feel: "smooth", "bouncy", "dramatic", "minimal"

Iterate

Don't try to say everything at once. Get the shape of the scene, then refine: "move the subtitle down", "make the entrance slower", "add a second shape on the right".

Tips and Limits

• Complex, multi-scene animations often come out best across a few messages rather than one giant prompt.
• Results vary — if something isn't right, ask for a change or edit it by hand.
• The assistant can generate image assets, but it doesn't import your own external files; upload images and 3D models yourself, then ask the assistant to arrange or animate them.

Next Steps

• Design Mode — Edit anything the assistant created
• Animation & Timeline — Fine-tune keyframes and easing
• Components — Turn AI output into reusable pieces
• Export & Publishing — Render your animation