Node Overview
Every "block" on the Martini canvas is a node. Different types of nodes have different capabilities, and by connecting them together you can build powerful AI workflows.
Node Types
| Node | Icon | Purpose |
|---|---|---|
| Text | Write prompts, text content | |
| Image | AI image generation and editing | |
| Video | Image-to-video, text-to-video | |
| Audio | Speech synthesis, sound effects, music | |
| 3D | 3D model generation | |
| Tool | Image upscaling, repair, lipsync, and more | |
| Note | Notes and annotations (doesn't participate in workflows) |
Basic Node Structure
Each node typically includes:
- Title: Node name, can be customized
- Input ports (left side): Receive content from other nodes
- Output ports (right side): Pass results to downstream nodes
- Settings area: Configure parameters, select models, etc.
- Action buttons: Generate, download, delete, etc.
Connection Rules
Not all nodes can be connected to each other. Common valid connections:
| Upstream Node | → | Downstream Node | Description |
|---|---|---|---|
| Text | → | Image | Use prompt to generate images |
| Text | → | Video | Use prompt to generate videos |
| Image | → | Video | Use image as first frame to generate video |
| Image | → | Audio | Generate audio description based on image content |
| Image | → | Tool | Upscale/repair images |
| Video | → | Video | Video editing |
| Video | → | Audio | Extract/generate audio from video |
| Video | → | Tool | Upscale/process video |
| Tool | → | Image | Processed image |
| Tool | → | Video | Processed video |
| Tool | → | Tool | Chain tool processing |
Next Steps
Choose a node you're interested in to learn more:
