Demeterics ChatInstall

Summary

Send a chat completion or response request to the Demeterics unified API using a specified large language model provider, model, and customizable options, receiving generated text along with metadata and possible tool calls.

Inputs

• API Endpoint (required) — Select the API endpoint to call, either the OpenAI-compatible chat completions or the Demeterics responses endpoint with extended features.
• Provider (required) — Choose the LLM provider from Groq, OpenAI, Anthropic, Google, or OpenRouter.
• Model (required) — Pick the specific model for chat completion, loaded dynamically based on the chosen provider.
• System Prompt — Optional system message to set or guide the model's behavior during the chat.
• User Message (required) — The user's input message to send to the model for generating a response.
• Options — Configure advanced parameters like temperature, max tokens, penalties, response format (text or JSON schemas), reasoning settings, web search options, service tiers, documents for retrieval-augmented generation, function-calling tools, tool usage controls, timeout, stop sequences, and other model-specific options.

Output shape

A list of JSON objects, one per input item, each containing chat response content, tool call data if any, model and provider identifiers, API endpoint used, usage statistics if provided, and the full original API response.

The output JSON has keys: content (the generated text), toolCalls (array of tool call blocks if any), model, provider, apiEndpoint, usage (token usage and cost data if provided), and fullResponse (entire API response). The response format can be plain text or structured JSON depending on input options. The node supports parallel execution over multiple input items, returning a list of results accordingly.

Links

• n8n community nodes installation guide