Overview
This node, named "Kirha Chat," provides access to a chat completion API specialized in delivering real-time data and insights about cryptocurrencies. It leverages language models to generate text completions based on user prompts, optionally enhanced with system-level instructions and configurable options.
Common scenarios where this node is beneficial include:
- Generating cryptocurrency market summaries or analysis.
- Creating automated responses or reports related to crypto trends.
- Integrating AI-driven chat capabilities into workflows that require up-to-date crypto information.
For example, you could use this node to prompt the model with "What are the latest trends in Bitcoin trading?" and receive a detailed, AI-generated response summarizing recent market activity.
Properties
| Name | Meaning |
|---|---|
| Model | The language model used to generate the completion. Options are dynamically loaded from the Kirha API and represent different available models. Default is "kirha:crypto". More info: Models and Limits. |
| System Prompt | An optional system-level prompt injected at the start of the chat to guide the model's behavior or context. This can be used to set tone, style, or specific instructions for the conversation. |
| Options | Additional configuration options for the completion request. Currently supports: • Sampling Temperature: Controls randomness of output (0 = deterministic, 1 = very random). Default is 0.5. |
| Prompt | The main user prompt injected by the Agent Node. This is the core input text for which the model generates a response. Supports expressions and is required. |
Output
The node outputs an array of JSON objects, each containing a result field with the generated text completion from the model.
Example output structure:
{
"result": "AI-generated response text based on the prompt"
}
No binary data is produced by this node.
Dependencies
- Requires an API key credential for authenticating with the Kirha API.
- Connects to the Kirha chat completion endpoint at
https://api.kirha.ai/chat/v1/openai. - The node dynamically fetches available models from the Kirha API before execution.
Troubleshooting
- Authentication errors: Ensure the API key credential is correctly configured and valid.
- Model loading issues: If the list of models fails to load, check network connectivity and API availability.
- Empty or unexpected responses: Verify that the prompt and system prompt are properly set and that the selected model supports the requested task.
- Temperature setting too high or low: Very low temperature values may produce repetitive outputs; very high values may produce incoherent results. Adjust accordingly.
Links and References
- Kirha API Documentation: Chat Completion OpenAI
- Models and Limits: Kirha Models