Venice Embeddings (Beta) icon

Venice Embeddings (Beta)

Generate vector embeddings from text with Venice.ai (Beta feature, limited access)
Join our community

Overview

The Venice Embeddings node generates vector embeddings from text using the Venice.ai API. It is designed to convert either a single text string or multiple text strings into numerical vector representations (embeddings), which are useful in various AI and machine learning applications such as semantic search, recommendation systems, clustering, and natural language understanding.

This node is currently in beta and requires special access permissions from Venice.ai. Typical use cases include:

  • Transforming user input or documents into embeddings for similarity comparison.
  • Preparing textual data for downstream machine learning models.
  • Generating embeddings for multiple texts in batch to speed up processing.

Properties

Name Meaning
BETA FEATURE - Limited Access Notice that this feature is in beta and requires special access permissions from Venice.ai.
Input Type How to process the input: "Single Text" (one string) or "Multiple Texts" (JSON array of strings).
Text The single text string to generate embeddings for (required if Input Type is Single Text).
Texts JSON array of text strings to generate embeddings for (required if Input Type is Multiple Texts).
Model The embedding model to use; currently only "BGE-M3" is available.
Options Additional options including:
- Dimensions Number of dimensions for the output embeddings (default 1024).
- Encoding Format Format of returned embeddings: "Float" (floating point numbers) or "Base64" (base64-encoded string).

Output

The node outputs an array of JSON objects, each containing the generated embeddings corresponding to the input texts. The structure depends on the encoding format selected:

  • If "Float" encoding is chosen, embeddings are arrays of floating-point numbers representing the vector.
  • If "Base64" encoding is chosen, embeddings are base64-encoded strings representing the vector.

Each output item corresponds to one input text (or the single input text) and includes metadata linking it back to the original input item index.

If errors occur during execution, error details are included in the output JSON under an error field, especially when "Continue On Fail" is enabled.

Dependencies

  • Requires an active API key credential for authenticating with the Venice.ai API.
  • The node makes HTTP POST requests to https://api.venice.ai/api/v1/embeddings.
  • The user must have beta access permissions granted by Venice.ai to use this embeddings feature.

Troubleshooting

  • Authentication Errors (401 Unauthorized):
    This node requires special beta access. If you receive authentication errors, contact Venice.ai to request beta access. The node will provide a clear error message indicating this requirement.

  • Invalid Input Format:
    When using multiple texts input, the node expects a valid JSON array of strings. Parsing errors or invalid formats will cause the node to throw an error specifying the issue.

  • Continue On Fail Behavior:
    If enabled, the node will continue processing other items even if some fail, returning error details in the output JSON for failed items.

Links and References

Discussion