Skip to main content
Welcome to the AI Canvas! This is your creative workspace where you can spawn nodes, connect different AI models, and create your own personalized production workflows using our node-based creative engine. Whether you’re generating images, manipulating video, or directing complex text prompts, the Canvas lets you build the perfect pipeline for your ideas.

Getting Started

Building on the canvas is designed to be intuitive and modular. You can start creating your workflow by combining different blocks:
  • Right-click anywhere on the canvas background to open the Canvas Menu and spawn new nodes (Image, Video, or Utility).
  • Drag and drop files directly from your computer or asset library to instantly create an Attachment node.
  • Right-click a node to access its specific Configuration Menu (Model selection, Size, Aspect Ratio).

Node & Port Specifications

Continuum uses a Typed Port System to ensure compatibility between nodes. Connections are visually color-coded based on their data type.
Node TypePrimary Inputs (Handles)Primary OutputPort Color
PromptNone (Manual Entry)textGrey (#6b7280)
NegativeNone (Manual Entry)textGrey (#6b7280)
AttachmentFile Upload / Library Dropimage / videoBlue / Purple
Nano Genprompt, ref-images, aspect, triggerimageBlue (#3b82f6)
Video Genprompt-in, negative, ref-images, first-frame, last-framevideoPurple (#a855f7)
LLM NodeuserPrompt, images, video, audio, documenttextGrey (#6b7280)
ModelNone (UI Selection)providerOrange (#f59e0b)
AspectNone (UI Selection)aspectGreen (#22c55e)
ArrayList of itemstext (stream)Grey (#6b7280)

Aspect Ratio Constraints

Aspect ratios are strictly validated against provider capabilities. The system automatically snaps node dimensions to these ratios while maintaining a minimum area.
Model ProviderSupported Aspect Ratios
Nano Banana1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9
Veo 3.116:9, 9:16

System Limitations & Configuration

These limits are enforced at the UI, validation, and execution levels to ensure high-density monitoring and reliability.

Execution Limits

  • Concurrency: Maximum 3 nodes can execute simultaneously across your canvas.
  • Dependency Resolution: A node remains in idle or pending state until all required incoming edges provide valid data.

Media & Content Limits

  • Reference Images:
    • nano-banana (Flash): Max 3 images.
    • nano-banana-pro: Max 14 images.
  • Video Duration: Fixed presets of 4s, 6s, or 8s for Veo models.
  • Image Resolution: Supports output sizing of 512px, 1K, 2K, and 4K.

Context Menu Breakdown

The AI Canvas features a contextual interaction system with two primary layers:

1. Canvas Context Menu (Background / Pane)

Right-click any empty area on the canvas to build and manage workflows.
SectionOptionShortcutPurpose
Add NodeImage, Video, Utility-Spawn specific generators or reference nodes.
WorkflowLoad Workflow-Opens the library of saved workflow templates.
ViewPan ModeHEnables “Hand” tool for navigating the canvas.
ViewSelect ModeVEnables box-selection for nodes and edges.
ViewFit ViewShift+FCenters and scales the canvas to show all nodes.
DangerClear Canvas-Wipes the entire graph (Destructive).

2. Node Context Menu (Common Items)

Right-click any node to access administrative and generative actions:
  • Duplicate (Cmd+D): Clones the node with its current data and configuration.
  • Delete (Del / Backspace): Removes the node and breaks its existing connections.
  • Run Node (R): (Generators only) Manually triggers the execution of that specific block.
  • Download Output: (Generators/Media only) Saves the generated artifact to your machine.

3. Specific Node Configurations

Generators expose deep configuration settings directly within their context menus:
  • Nano Gen: Toggle between Flash, Pro, and Nano Banana 2 models. Set output resolution and aspect ratio (1:1, 16:9, 9:16, 4:3, 3:4).
  • Veo 3.1: Select between standard and Fast models. Toggle between Cinematic (16:9) and Vertical (9:16) aspect ratios.
  • LLM Node: Use Enrich Prompt (R) to expand simple text into high-fidelity instructions.

Workflow Logic & UX

  • Real-time Streaming: LLM nodes use Server-Sent Events (SSE) for real-time text delivery.
  • Port Validation: The system prevents illegal connections (e.g., connecting a video output to an aspect input) via real-time port validation.
  • Status Indicators: Nodes display visual status badges: idle (gray), running (pulse), success (green), and error (red).
  • Anchor Point: Nodes are spawned exactly where your mouse was clicked on the canvas.
  • Contextual Guardrails: The menu dynamically disables options (e.g., “Download”) if the node hasn’t successfully generated an artifact yet.

Supported Models

Continuum supports a range of powerful AI models to suit different creative needs:
ModelDescription
Nano Banana 2Our latest high-performance model for creative generation.
Nano BananaReliable and versatile generation.
Nano Banana FastOptimized for speed and rapid iteration.
Veo 3.1Advanced image and video synthesis.
Veo 3.1 FastHigh-speed Veo generation.
We are constantly exploring new model integrations based on user feedback to ensure you have access to the best creative tools.

Pro Tips

Smart Previews: Media nodes automatically preserve your source’s professional aspect ratio.
  • Collaborate: Invite your team! You can see their cursors and work together in real-time.
  • Extendable Workflow: The canvas is built on a modular system of nodes and edges, allowing you to easily expand and reorganize your creative pipelines as your project grows.