Tovaki
Home
Analyze
Canvas
Editor
POD Studio
MCP Connect
Upgrade
Docs
Tovaki
Docs
Getting started
First steps: one-line creation and core concepts
Canvas
Everyday canvas, cards and toolbar operations
AI creation
Director chat plus image and video generation
Advanced
Skills, model choice, export and sharing
Updates
Product changelog
All articles
Quick start
Type one creative idea into the home composer (e.g. "a cozy 15s vertical short about hand-brewed coffee"). The AI director first offers 2-3 distinct creative directions; pick one and it splits the storyboard, writes every shot prompt, and lays the shot cards out on the canvas.
How Tovaki works
Tovaki = an infinite canvas + an AI director you talk to. You describe what you want in natural language; the director plans, the canvas holds every generated asset.
Canvas & cards
The canvas sits on a dot grid; changes autosave in real time (a "Saved" indicator sits top-left). Drag on empty space to box-select: every card the box touches gets selected. Pan by holding Space or the middle mouse button and dragging, or by swiping two fingers on a trackpad; zoom with a pinch or ⌘/Ctrl + scroll. On phones and tablets, drag with one finger to pan and pinch to zoom.
Toolbar & shortcuts
"Add node" creates video / image / text / audio cards; "Project ratio" switches 9:16 / 16:9 / 21:9 and new cards follow it by default.
The AI director
The right panel is the AI director. Say your idea → it returns 2-3 creative directions (core idea / story arc / emotional curve / climax / duration / differentiation); storyboarding only starts after you pick one, and that stage is free.
Generating images
Two ways: ① tell the director in Image mode and it writes the prompt and drops an image card; ② "Add node → Image" to build one manually.
Generating video
Two ways: ① say the idea to the director and it splits the storyboard into shot cards; ② "Add node → Video" for an empty shot card.
Skills
Clicking a skill doesn't open a form — it docks as a chip above the composer and switches the guidance to that skill's workflow; click again to remove.
Choosing models
Two levels: ① the card's model dropdown affects that card; ② the director's model preference decides the default written into new shot cards.
Export & sharing
Single asset: right-click a card → download at original resolution; the "Generated files" panel downloads per item too.
Changelog
2026-10-08 · Veo 3.1 is retired; Gemini Omni 1.1 Flash takes its place
Docs
Tovaki
Docs
Getting started
First steps: one-line creation and core concepts
Canvas
Everyday canvas, cards and toolbar operations
AI creation
Director chat plus image and video generation
Advanced
Skills, model choice, export and sharing
Updates
Product changelog
All articles
Quick start
Type one creative idea into the home composer (e.g. "a cozy 15s vertical short about hand-brewed coffee"). The AI director first offers 2-3 distinct creative directions; pick one and it splits the storyboard, writes every shot prompt, and lays the shot cards out on the canvas.
How Tovaki works
Tovaki = an infinite canvas + an AI director you talk to. You describe what you want in natural language; the director plans, the canvas holds every generated asset.
Canvas & cards
The canvas sits on a dot grid; changes autosave in real time (a "Saved" indicator sits top-left). Drag on empty space to box-select: every card the box touches gets selected. Pan by holding Space or the middle mouse button and dragging, or by swiping two fingers on a trackpad; zoom with a pinch or ⌘/Ctrl + scroll. On phones and tablets, drag with one finger to pan and pinch to zoom.
Toolbar & shortcuts
"Add node" creates video / image / text / audio cards; "Project ratio" switches 9:16 / 16:9 / 21:9 and new cards follow it by default.
The AI director
The right panel is the AI director. Say your idea → it returns 2-3 creative directions (core idea / story arc / emotional curve / climax / duration / differentiation); storyboarding only starts after you pick one, and that stage is free.
Generating images
Two ways: ① tell the director in Image mode and it writes the prompt and drops an image card; ② "Add node → Image" to build one manually.
Generating video
Two ways: ① say the idea to the director and it splits the storyboard into shot cards; ② "Add node → Video" for an empty shot card.
Skills
Clicking a skill doesn't open a form — it docks as a chip above the composer and switches the guidance to that skill's workflow; click again to remove.
Choosing models
Two levels: ① the card's model dropdown affects that card; ② the director's model preference decides the default written into new shot cards.
Export & sharing
Single asset: right-click a card → download at original resolution; the "Generated files" panel downloads per item too.
Changelog
2026-10-08 · Veo 3.1 is retired; Gemini Omni 1.1 Flash takes its place