generated: '2026-07-19'
method: derived
source: openapi/flexai-token-factory-openapi.yml
description: >-
  Entity graph derived from the FlexAI Token Factory OpenAPI schemas. The API is
  stateless request/response inference (OpenAI-compatible); the only persistent
  resource is the Model catalog. Chat completions are transient response objects
  composed of messages and choices.
entities:
  - name: Model
    id_field: id
    description: A served open model (text, code, reasoning, vision, embedding, image, video, or audio).
    source_schema: '#/components/schemas/Model'
  - name: ChatCompletion
    id_field: id
    description: A transient chat completion response.
    source_schema: '#/components/schemas/ChatCompletion'
  - name: ChatMessage
    description: A single message in a conversation (role + content).
    source_schema: '#/components/schemas/ChatMessage'
  - name: Choice
    description: One generated choice within a chat completion.
    source_schema: '#/components/schemas/ChatCompletion/choices'
  - name: Usage
    description: Token accounting for a completion (prompt/completion/total tokens).
    source_schema: '#/components/schemas/ChatCompletion/usage'
relationships:
  - from: ChatCompletion
    to: Choice
    kind: has_many
    via: choices
  - from: Choice
    to: ChatMessage
    kind: has_one
    via: message
  - from: ChatCompletion
    to: Usage
    kind: has_one
    via: usage
  - from: ChatCompletion
    to: Model
    kind: belongs_to
    via: model
  - from: ChatCompletionRequest
    to: ChatMessage
    kind: has_many
    via: messages