Text Node

A Text Node holds a block of text. You can type the text yourself, or use an AI language model to generate it. The text can then flow downstream into Image, Video, or Audio Nodes as part of their prompt.

Adding content

Typing your own text

When a Text Node is empty, the display area shows a "Enter your own content" preset. Click it to turn the display area into a text input field — type or paste up to 5,000 characters.

Your typed content and the generation panel coexist. If you later use the model to generate new text, the generated result replaces whatever you typed.

Generating with a model

Write a prompt in the generation panel's input field and click Generate. The model returns text that appears directly in the display area, preserving any Markdown formatting in the response.

If the input field is empty and no upstream text nodes are connected, the Generate button is disabled.

Tip: You can leave the prompt field empty if an upstream Text Node is connected — the upstream text becomes the full prompt.

Generation panel

Element	Description
Reference strip	Shows thumbnails of connected upstream Image and Video nodes
Prompt input	Where you type your instructions; up to 5,000 characters
Upstream text icons	Inline icon blocks representing connected upstream Text Nodes
Model selector	Choose the language model for this generation
Generate button	Shows credit cost; click to confirm and start generation

Referencing upstream content in your prompt

When upstream Image or Video nodes are connected, they appear in the reference strip numbered in connection order. Use @Image 1, @Image 2, @Video 1, etc. in your prompt to direct the model's attention to a specific asset.

Upstream text from connected Text Nodes appears as inline icon blocks inside the prompt field itself, not in the reference strip. The block represents the upstream text being prepended to your own input when the request is sent to the model.

When the Generate button is disabled

Generation is already in progress on this node
The prompt field is empty and no upstream text is connected
An upstream node is currently generating ("Upstream task in progress.")

Available models

The default model is Gemini 3.1-Pro. For the full list of text models, their context limits, and costs, see Text Models.

Connecting to other nodes

Accepted upstream connections

Upstream node	How its content is used
Text Node	Content becomes an inline icon block in the prompt field, prepended to your input
Image Node	Image appears in the reference strip; reference with `@Image N` in your prompt
Video Node	Video appears in the reference strip; reference with `@Video N` in your prompt

Supported downstream connections

A Text Node can connect downstream to a Text Node, Image Node, Video Node, or Audio Node. Its text content will appear as an inline icon block in the downstream node's prompt field.

Display area toolbar

When the display area contains text, a toolbar appears with:

Download — saves the content as a .txt file

Tip: The Text Node display area does not support in-place editing once content is generated. To revise the text, update your prompt and regenerate, or delete the node and start fresh.

Content and deletion

Deleting a Text Node discards its content — text is not saved to your account assets. If you need to keep the text, download it before deleting the node.