VeniceAi

Use VeniceAi AI models.

Actions4

Overview

The VeniceAi node's "Images" operation allows you to generate images using Venice AI models based on a text prompt and a variety of customizable options. This is useful for automating image creation in workflows, such as generating illustrations, concept art, or visual assets from textual descriptions. Practical scenarios include content generation for marketing, prototyping designs, or creating unique visuals for social media posts.

Properties

Name	Type	Meaning
Model Name or ID	options	Select the AI model to use for image generation. You can choose from a list or specify an ID via expression.
Prompt	string	The main text prompt describing what image to generate.
Image Options	collection	A group of advanced settings for image generation (see below for sub-options).

Image Options (collection):

Name	Type	Meaning
CFG Scale	number	Controls adherence to the prompt; higher values mean more faithful results.
Embed EXIF Metadata	boolean	If enabled, embeds generation metadata into the image's EXIF data.
Format	options	Output image format: PNG or WEBP.
Width	number	Width of the generated image (must be divisible by a model-specific divisor).
Height	number	Height of the generated image (must be divisible by a model-specific divisor).
Hide Watermark	boolean	Optionally removes the Venice watermark from the image.
LoRA Strength	number	Adjusts LoRA strength if the model supports additional LoRAs.
Negative Prompt	string	Specifies elements to exclude from the generated image.
Return Binary	boolean	If true, returns binary image data instead of base64-encoded data.
Safe Mode	boolean	Blurs images classified as containing adult content.
Seed	number	Sets the random seed for reproducibility; 0 means a new random seed each time.
Steps	number	Number of inference steps; higher values may improve quality but increase processing time.
Style Preset	options	Choose a style preset (e.g., Photographic, Anime, Digital Art, etc.).
Sampler	string	Sampling method used during image generation.
Inpaint	collection	Options for inpainting (modifying parts of an existing image).

Inpaint (sub-collection):

Name	Type	Meaning
Strength	number	Degree of inpainting effect applied.
Source Image	string	Base64-encoded source image to inpaint.
Mask Options	collection	Further options for specifying which part of the image to inpaint and how.

Mask Options (sub-collection):

Name	Type	Meaning
Image Prompt	string	Text prompt describing the original input image for inpainting.
Object Target	string	Elements in the original image to be replaced/overwritten.
Inferred Object	string	Content to add to the image via inpainting, replacing the object target.

Output

If Return Binary is enabled:
- The output will contain a binary property with the generated image file (typically under the key data). The file will have the specified format (PNG or WEBP).
- The json property will be present but may be empty or contain minimal metadata.
If Return Binary is disabled:
- The output will contain a json.response field with the API response, which typically includes a base64-encoded image and possibly other metadata.

Example Output (non-binary):

{
  "response": {
    "image_base64": "<base64-string>",
    "metadata": { /* ... */ }
  }
}

Example Output (binary):

{
  "binary": {
    "data": {
      "fileName": "image.png",
      "mimeType": "image/png",
      "data": "<binary-data>"
    }
  },
  "json": {}
}

Dependencies

External Service: Requires access to the Venice AI API.
API Key: You must provide valid Venice AI API credentials (apiKey and baseUrl) in n8n.
n8n Configuration: The node expects the Venice AI API credentials to be set up in the n8n credential store as veniceAiApi.

Troubleshooting

No valid API key provided: Ensure your Venice AI API credentials are correctly configured in n8n.
Invalid response format from Venice AI API: Double-check your input parameters, especially the model selection and prompt formatting.
No models found for operation: Make sure your API key has access to the required models and that the correct operation is selected.
Width/Height errors: Some models require image dimensions to be divisible by a specific value. Check the model documentation if you encounter dimension-related errors.
Binary data issues: If enabling "Return Binary," ensure downstream nodes can handle binary data.