RahyanaInstall

Summary

Generate chatbot completions using a specified Rahyana AI model by sending a user message and optional system prompt, returning the assistant's response text or detailed JSON including metadata.

Inputs

• Model (required) — Model slug to use for the AI completion, such as 'google/gemini-2.0-flash-lite-001'.
• Chat Input (required) — User's chat message input; supports expressions for dynamic input.
• System Prompt — Optional system instruction to guide the AI's behavior.
• Temperature — Controls randomness of output; 0 is deterministic, 2 is very random.
• Max Tokens — Maximum number of tokens for the generated response.
• Stream — Enable streaming response mode (currently not differentiated in output handling).
• Return Raw — Return the full raw API response instead of only the assistant's content.
• Output Mode — Controls output format: 'content' returns the assistant's text response only, 'json' returns content plus model metadata.

Output shape

a single JSON object per execution, either containing the assistant's text response, a JSON with response content and model metadata, or the full raw API response

If 'Return Raw' is enabled, the full API response is emitted. Otherwise, the output depends on 'Output Mode': 'content' outputs a string response; 'json' outputs an object with 'content' and 'model'. Errors can appear in output JSON if 'Continue On Fail' is enabled.