Arcee AI Completions API

openapi: 3.2.0
info:
  title: AFM Completions API
  version: 0.1.0
tags:
- name: Completions
paths:
  /app/v1/completions/{session_id}/stream:
    post:
      tags:
      - Completions
      summary: Stream Message Completion
      description: Send a message and get a streaming completion response.
      operationId: stream_message_completion_app_v1_completions__session_id__stream_post
      security:
      - HTTPBearer: []
      parameters:
      - name: session_id
        in: path
        required: true
        schema:
          type: string
          format: uuid
          title: Session Id
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatInput'
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
  /app/v1/completions/create-chat:
    post:
      tags:
      - Completions
      summary: Create Chat With First Message
      description: 'Create new chat and stream first message response in one request.

        Optimized: streams immediately, saves to DB asynchronously at the end.'
      operationId: create_chat_with_first_message_app_v1_completions_create_chat_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateChatWithMessageRequest'
        required: true
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - HTTPBearer: []
  /app/v1/completions/{session_id}/regenerate:
    post:
      tags:
      - Completions
      summary: Regenerate Last Response
      description: 'Regenerate the last assistant response for a chat session.


        This endpoint:

        1. Gets the last user message and its paired assistant response

        2. Deletes only the assistant message (preserving the user message and its files)

        3. Streams a new response using the same user message content and settings'
      operationId: regenerate_last_response_app_v1_completions__session_id__regenerate_post
      security:
      - HTTPBearer: []
      parameters:
      - name: session_id
        in: path
        required: true
        schema:
          type: string
          format: uuid
          title: Session Id
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
  /app/v1/completions/messages/{message_id}/stop:
    post:
      tags:
      - Completions
      summary: Stop Message Stream
      description: 'Signal the backend to stop an in-progress stream for the given assistant message.

        Works across pods by setting a Valkey key that the streaming generator polls per chunk.'
      operationId: stop_message_stream_app_v1_completions_messages__message_id__stop_post
      security:
      - HTTPBearer: []
      parameters:
      - name: message_id
        in: path
        required: true
        schema:
          type: string
          format: uuid
          title: Message Id
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
  /app/v1/completions/generate-title:
    post:
      tags:
      - Completions
      summary: Generate Chat Title
      description: Generate a title for a chat session based on the initial message.
      operationId: generate_chat_title_app_v1_completions_generate_title_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/GenerateTitleRequest'
        required: true
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/GenerateTitleResponse'
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
      security:
      - HTTPBearer: []
  /api/v1/chat/completions:
    post:
      tags:
      - Completions
      summary: Create Chat Completion
      description: Chat completion endpoint using the shared LiteLLM-backed external service.
      operationId: create_chat_completion_api_v1_chat_completions_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatCompletionRequest'
        required: true
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
  /v1/chat/completions:
    post:
      tags:
      - Completions
      summary: Create Chat Completion
      description: Chat completion endpoint using the shared LiteLLM-backed external service.
      operationId: create_chat_completion_v1_chat_completions_post
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatCompletionRequest'
        required: true
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema: {}
        '422':
          description: Validation Error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/HTTPValidationError'
components:
  schemas:
    ChatCompletionRequest:
      properties:
        model:
          type: string
          minLength: 1
          title: Model
        messages:
          items: {}
          type: array
          minItems: 1
          title: Messages
        stream:
          type: boolean
          title: Stream
          default: false
        temperature:
          title: Temperature
        top_p:
          title: Top P
        top_k:
          title: Top K
        min_p:
          title: Min P
        n:
          anyOf:
          - type: integer
            minimum: 1.0
          - type: 'null'
          title: N
          description: Kept structural because response accounting requires the expected choice count.
        stop:
          title: Stop
        seed:
          title: Seed
        max_tokens:
          title: Max Tokens
        max_completion_tokens:
          title: Max Completion Tokens
        presence_penalty:
          title: Presence Penalty
        frequency_penalty:
          title: Frequency Penalty
        logit_bias:
          title: Logit Bias
        logprobs:
          title: Logprobs
        top_logprobs:
          title: Top Logprobs
        tools:
          title: Tools
        tool_choice:
          title: Tool Choice
        parallel_tool_calls:
          title: Parallel Tool Calls
        functions:
          title: Functions
        function_call:
          title: Function Call
        reasoning_effort:
          title: Reasoning Effort
        reasoning:
          title: Reasoning
        response_format:
          title: Response Format
        stream_options:
          title: Stream Options
        user:
          title: User
        metadata:
          title: Metadata
        cache_control:
          title: Cache Control
        image_config:
          title: Image Config
        modalities:
          title: Modalities
        prediction:
          title: Prediction
        prompt_cache_key:
          title: Prompt Cache Key
        prompt_cache_options:
          title: Prompt Cache Options
        repetition_penalty:
          title: Repetition Penalty
        top_a:
          title: Top A
        trace:
          title: Trace
      additionalProperties: false
      type: object
      required:
      - model
      - messages
      title: ChatCompletionRequest
      description: 'Supported public OpenAI-compatible request fields.


        This is deliberately a positive allowlist. LiteLLM mixes user-facing

        fields with credentials, URLs, retry knobs, fallback controls, and testing

        hooks in one loose kwargs surface; accepting arbitrary extras would let a

        caller take ownership of server-side routing and transport policy. Values

        inside the safe inference allowlist intentionally remain untyped so the

        selected provider owns request validation.'
    CreateChatWithMessageRequest:
      properties:
        message:
          type: string
          maxLength: 50000
          title: Message
          description: First message content
        title:
          anyOf:
          - type: string
            maxLength: 600
          - type: 'null'
          title: Title
          description: Optional chat title override
        base_model_name:
          anyOf:
          - type: string
          - type: 'null'
          title: Base Model Name
          description: Base model name (e.g., 'trinity-mini-26b', 'gpt-4o')
        chat_id:
          anyOf:
          - type: string
          - type: 'null'
          title: Chat Id
          description: Pre-generated chat ID (for file upload path consistency)
        files:
          anyOf:
          - items:
              additionalProperties: true
              type: object
            type: array
          - type: 'null'
          title: Files
        fileReferences:
          anyOf:
          - items:
              additionalProperties: true
              type: object
            type: array
          - type: 'null'
          title: Filereferences
        tool:
          anyOf:
          - type: string
          - type: 'null'
          title: Tool
        systemPrompt:
          anyOf:
          - additionalProperties: true
            type: object
          - type: 'null'
          title: Systemprompt
          description: System prompt object for the chat
        enabledTools:
          anyOf:
          - items:
              type: string
            type: array
          - type: 'null'
          title: Enabledtools
        temperature:
          anyOf:
          - type: number
          - type: 'null'
          title: Temperature
          description: Model temperature
        top_p:
          anyOf:
          - type: number
          - type: 'null'
          title: Top P
          description: Top-p (nucleus) sampling
        top_k:
          anyOf:
          - type: integer
          - type: 'null'
          title: Top K
          description: Top-k sampling
        min_p:
          anyOf:
          - type: number
          - type: 'null'
          title: Min P
          description: Min-p sampling threshold
        provider_preference:
          anyOf:
          - type: string
          - type: 'null'
          title: Provider Preference
          description: Preferred provider name for model selection
      type: object
      required:
      - message
      title: CreateChatWithMessageRequest
      description: Model for creating chat and sending first message in one request.
    HTTPValidationError:
      properties:
        detail:
          items:
            $ref: '#/components/schemas/ValidationError'
          type: array
          title: Detail
      type: object
      title: HTTPValidationError
    ValidationError:
      properties:
        loc:
          items:
            anyOf:
            - type: string
            - type: integer
          type: array
          title: Location
        msg:
          type: string
          title: Message
        type:
          type: string
          title: Error Type
      type: object
      required:
      - loc
      - msg
      - type
      title: ValidationError
    GenerateTitleResponse:
      properties:
        title:
          type: string
          title: Title
          description: Generated chat title
        chat_session_id:
          type: string
          format: uuid
          title: Chat Session Id
          description: Chat session ID that was updated
      type: object
      required:
      - title
      - chat_session_id
      title: GenerateTitleResponse
      description: Response with generated title.
    ChatInput:
      properties:
        content:
          type: string
          maxLength: 50000
          title: Content
          description: Message content (max 50,000 characters)
        session_id:
          type: string
          format: uuid
          title: Session Id
        data_group_id:
          anyOf:
          - type: string
          - type: 'null'
          title: Data Group Id
        system_prompt:
          anyOf:
          - type: string
          - additionalProperties: true
            type: object
          - type: 'null'
          title: System Prompt
          description: System prompt (string or object)
        meta:
          anyOf:
          - additionalProperties: true
            type: object
          - type: 'null'
          title: Meta
        tool:
          anyOf:
          - type: string
          - type: 'null'
          title: Tool
          description: Tool to use (e.g., 'search')
        provider_preference:
          anyOf:
          - type: string
          - type: 'null'
          title: Provider Preference
          description: Preferred provider name for model selection
      type: object
      required:
      - content
      - session_id
      title: ChatInput
    GenerateTitleRequest:
      properties:
        initial_message:
          type: string
          maxLength: 50000
          title: Initial Message
          description: Initial message to generate title from
        chat_session_id:
          type: string
          format: uuid
          title: Chat Session Id
          description: Chat session ID to update with generated title
      type: object
      required:
      - initial_message
      - chat_session_id
      title: GenerateTitleRequest
      description: Request to generate a chat title. The backend uses a pinned title-generation model.
  securitySchemes:
    HTTPBearer:
      type: http
      scheme: bearer
x-provenance:
  method: harvested
  publisher: Arcee AI
  source: https://api.arcee.ai/openapi.json
  note: split by refine-openapis from the provider's own published document, verified by fetch (openapi/_original/*.fetch.json);
    stamped by stamp-openapi-provenance.py (roadmap#793)