openai_chat: Send LLM Messages to the OpenAI Responses API

Description

Sends a message history to the OpenAI Responses API (POST /v1/responses) and returns the assistant's reply. Supports streaming, tool use, structured output, and reasoning models (o-series) via .reasoning_effort.

Usage

openai_chat(
  .llm,
  .model = "gpt-5.6-terra",
  .max_output_tokens = NULL,
  .temperature = NULL,
  .seed = NULL,
  .stream = FALSE,
  .timeout = 60,
  .verbose = FALSE,
  .json_schema = NULL,
  .max_tries = 3,
  .dry_run = FALSE,
  .reasoning_effort = NULL,
  .tools = NULL,
  .tool_choice = NULL,
  .max_tool_rounds = 10,
  .stateful = FALSE
)

Value

A new LLMMessage object with the assistant's response appended.

Arguments

.llm: An LLMMessage object containing the conversation history.
.model: The model identifier (default: "gpt-5.6-terra").
.max_output_tokens: Maximum tokens to generate (caps reasoning + completion).
.temperature: Sampling temperature (0-2).
.seed: Seed for deterministic sampling.
.stream: If TRUE, stream output to console (default: FALSE).
.timeout: Request timeout in seconds (default: 60).
.verbose: Print rate-limit info after the call (default: FALSE).
.json_schema: A tidyllm schema or ellmer type for structured output.
.max_tries: Maximum retry attempts (default: 3).
.dry_run: If TRUE, return the request object without sending (default: FALSE).
.reasoning_effort: For o-series models: "low", "medium", or "high".
.tools: A TOOL object or list of TOOL objects for function calling.
.tool_choice: Tool selection behavior: "none", "auto", or "required".
.max_tool_rounds: Maximum tool-loop iterations (default: 10).
.stateful: If TRUE, use previous_response_id to pass conversation context server-side instead of re-sending full message history. Falls back to full-history mode silently when no previous response ID is available (first turn or prior turn used a different provider), and with a warning if the server rejects the stored ID (e.g., expired context). Default: FALSE.