Generate images from text prompts using AI. Supports multiple providers with varying quality, speed, and cost.

Generate Image

The Generate Image node creates images from text prompts using AI image generation models. It is one of the most commonly used nodes in Nodaro workflows, typically placed after a Generate Script or Text Prompt node.

Providers

Choose a provider in the config panel. Each offers different quality, speed, and credit cost trade-offs:

Provider	Credits	Best For
Nano Banana	4	Fast drafts, iteration, storyboards
Nano Banana Pro	6	Higher detail, production-ready images
Flux	10	Highest quality, photorealistic output
Grok	8	Creative and stylized imagery
GPT Image	12	Text rendering, complex compositions

Tip

Start with Nano Banana while prototyping your workflow to conserve credits. Switch to Flux or GPT Image for your final render.

Configuration

The config panel exposes these settings:

Setting	Description	Default
Provider	Which AI model to use	Nano Banana Pro
Aspect Ratio	Output dimensions (1:1, 16:9, 9:16, 4:3, 3:4)	16:9
Negative Prompt	What to exclude from the image	(empty)
Style Preset	Apply a predefined style (cinematic, anime, photographic, etc.)	None
Seed	Fixed seed for reproducible results	Random

Aspect Ratio

The aspect ratio determines the image dimensions. Common choices:

16:9 -- Landscape, standard video frame. Best for most video workflows.
9:16 -- Portrait, vertical video (TikTok, Reels, Shorts).
1:1 -- Square, social media posts.
4:3 -- Classic TV/photo ratio.

Negative Prompt

Use negative prompts to steer the model away from unwanted elements:

blurry, low quality, watermark, text overlay, distorted hands, extra fingers

Not all providers support negative prompts with equal effectiveness. Flux and Nano Banana Pro respond best to negative prompts.

Inputs

Handle	Type	Required	Description
prompt	Text	Yes	The text description of the image to generate. Can come from a Text Prompt node, Generate Script node, or any text-producing node.
reference	Image	No	One or more reference images for style or character consistency. Connect from Upload Image, Character, Object, or Location nodes.

Reference Images and Character Consistency

Connect a Character node or Upload Image node to the reference input to maintain visual consistency across multiple generated images. This is especially useful when:

Generating multiple scenes featuring the same character
Creating a series of product shots with consistent styling
Building storyboards where visual continuity matters

You can connect multiple reference sources. The AI model will attempt to blend the reference characteristics into the generated output.

Character Consistency

For best results with character consistency, use the AI Agent node. It generates multiple image prompts from a single concept and automatically connects reference images to each spawned Generate Image node.

Outputs

Handle	Type	Description
image	Image	The generated image. Connect to Image to Video, Image to Image, Edit Image, Merge Video & Audio, or Save to Storage.

The output image can connect to multiple downstream nodes simultaneously. For example, you can send the same image to both an Image to Video node and a Save to Storage node.

Parameter Inputs

These settings can also be driven by Parameter nodes connected to the Generate Image node:

Parameter Node	Controls
Image Provider	Switches the AI provider
Aspect Ratio	Sets the output aspect ratio
Text	Overrides or appends to the prompt

When a Parameter node is connected, its value overrides the config panel setting. This lets you build dynamic workflows where a single parameter change updates multiple nodes.

Usage in Workflows

Basic: Prompt to Image

Text Prompt --> Generate Image --> Save to Storage

Video Pipeline

Text Prompt --> Generate Script --> Generate Image --> Image to Video --> Combine Videos

Character-Consistent Multi-Scene

Character --------+
                  |
AI Agent --> Generate Image (Scene 1) --> Image to Video
         --> Generate Image (Scene 2) --> Image to Video
         --> Generate Image (Scene 3) --> Image to Video

With Style Reference

Upload Image (style ref) --+
                            |
Text Prompt ------------> Generate Image --> Image to Video

List Execution

When connected to a Loop node or receiving list input, Generate Image executes once per item. For example, if a Loop node provides 5 prompts, the node generates 5 separate images sequentially.

A progress badge shows the current iteration (e.g., "2/5") and completed images appear as expandable clones on the canvas.

Tips

Prompt quality matters. Detailed, specific prompts produce better results than vague ones. Include lighting, mood, camera angle, and style references in your prompt.
Use Generate Script first. For video workflows, let the Generate Script node create optimized image prompts from your concept rather than writing them manually.
Iterate cheaply. Use Nano Banana (4 credits) to test your workflow, then switch to Flux (10 credits) or GPT Image (12 credits) for final output.
Negative prompts help. If you see recurring artifacts, add them to the negative prompt.
Seed for consistency. Set a fixed seed when you want reproducible results across multiple runs.

Generate Image

On this page