# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Parallel Chat API (Beta) API
  version: 1.0.0
extends: openapi/parallel-chat-api-beta-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-10-01'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 1
- target: $.paths['/v1beta/chat/completions'].post
  update:
    x-apievangelist-phrasing:
      intent: Get a chat completion grounded in web research
      effect: write
      questions:
      - Can I get chat-style answers from Parallel using an OpenAI-compatible chat completions request?
      - Is it possible to stream chat completion tokens back as they are generated?
      - Which sampling settings like temperature or max_tokens does the chat completions endpoint actually honor?
      - Can a chat completion continue from an earlier interaction instead of resending the whole history?
      instructions:
      - text: Send the chat messages {messages} to the {model} model and return the completion.
        slots:
          messages: requestBody.messages
          model: requestBody.model
      - text: 'Stream a chat completion from {model} for these messages: {messages}.'
        slots:
          model: requestBody.model
          messages: requestBody.messages
      - text: Continue the chat from interaction {previous_interaction_id} with the new messages {messages} using {model}.
        slots:
          previous_interaction_id: requestBody.previous_interaction_id
          messages: requestBody.messages
          model: requestBody.model
      method: generated
      generated: '2026-10-01'