Overview
This node integrates with Venice.ai's chat completion models to generate text completions based on user prompts. It is designed for scenarios where users want to leverage advanced language models for tasks such as content generation, conversational AI, summarization, or any application requiring natural language understanding and generation.
Typical use cases include:
- Generating chatbot responses in automated workflows.
- Creating creative writing or marketing copy.
- Summarizing or expanding text inputs.
- Integrating AI-driven language capabilities into larger automation chains.
The node allows fine-tuning of the model's behavior through various parameters controlling randomness, token limits, penalties, and response formatting.
Properties
| Name | Meaning |
|---|---|
| Connect this to AI Chain or AI Agent nodes | A notice guiding users to connect this node with AI Chain or AI Agent nodes for extended functionality. |
| Model | The specific Venice.ai model to use for generating chat completions (e.g., "llama-3.3-70b"). |
| Options | Collection of parameters to customize the generation: |
| - Frequency Penalty | Penalizes new tokens based on their existing frequency (-2 to 2). |
| - Maximum Number of Tokens | Maximum number of tokens to generate in the output. |
| - Maximum Temperature | Upper bound for dynamic temperature scaling (0 to 2). |
| - Minimum P | Minimum probability threshold for token selection (0 to 1). |
| - Minimum Temperature | Lower bound for dynamic temperature scaling (0 to 2). |
| - Number of Completions | How many different completion choices to generate. |
| - Presence Penalty | Penalizes new tokens based on their presence in the generated text so far (-2 to 2). |
| - Repetition Penalty | Parameter to reduce repetition; 1.0 means no penalty (minimum 0). |
| - Seed | Random seed for reproducible outputs. |
| - Stop Sequences | String or JSON array of sequences where the API will stop generating further tokens. |
| - Temperature | Controls randomness of output; lower values produce more deterministic results (0 to 2). |
| - Top K | Number of highest probability tokens to keep for top-k filtering. |
| - Top P | Controls diversity via nucleus sampling (0 to 1). |
| - Venice Parameters | Additional Venice-specific options: |
| - Character Slug | Identifier for a public Venice character to influence style or persona. |
| - Enable Web Search | Whether to enable web search during generation ("auto", "on", "off"). |
| - Include Venice System Prompt | Whether to include Venice system prompts alongside user-specified system prompts. |
| - Response Format | Defines the format of the response: |
| - Type | Currently supports "JSON Object" format. |
| - JSON Schema | JSON Schema used to validate and format the response when using JSON schema type. |
Output
The node outputs data under the json field containing the generated chat completions from the Venice.ai model. The structure depends on the chosen response format but typically includes the generated text or structured JSON objects conforming to the provided JSON schema.
If binary data were supported, it would represent associated files or media generated by the model, but this node focuses on textual completions only.
Dependencies
- Requires an API key credential for Venice.ai to authenticate requests.
- Uses Venice.ai's REST API endpoint at
https://api.venice.ai/api/v1. - Relies on internal helper classes for handling API calls, tracing, and error handling.
- No additional external dependencies beyond the Venice.ai service and n8n environment.
Troubleshooting
- Authentication errors: Ensure the Venice.ai API key credential is correctly configured and valid.
- Invalid parameter values: Parameters like temperature, penalties, and token counts must be within specified ranges; otherwise, the API may reject the request.
- Malformed stop sequences: If providing stop sequences as JSON arrays, ensure proper JSON formatting.
- Response validation failures: When using JSON schema response formatting, invalid schemas or mismatched responses can cause errors.
- API rate limits or network issues: May result in failed attempts; the node includes retry handlers but persistent failures require checking API usage limits and connectivity.