JoinGonka

Interact with JoinGonka Gateway API

Actions2

Chat Completion

AI-generated

Summary

Send a chat completion request to the JoinGonka Gateway API using an optional specified model identifier and a series of messages in OpenAI format to generate AI-driven chat responses. If the model field is left empty, the gateway's current default model will be used, reflecting the dynamic set of models available on the network.

Inputs

  • Model — Model identifier from the List Models operation. Leave empty to use the gateway's current default model — the set of models on the network changes over time.
  • Messages (required) — An array of JSON-formatted chat messages, each containing a 'role' (system, user, or assistant) and 'content', representing the conversation history to prompt the AI.
  • Temperature — Sampling temperature to control randomness in the response; 0 for deterministic and up to 2 for more creative outputs. Defaults to 0.7.
  • Max Tokens — Maximum number of tokens to generate in the AI response. Defaults to 1024.

Output shape

a single JSON object per input item

Returns the full JSON response from the JoinGonka Gateway's /chat/completions endpoint containing the generated chat completion. If multiple input items are processed, the node outputs one corresponding response per item. Errors can be optionally captured per item if 'Continue On Fail' is enabled.

Links

Discussion