LM Studio OpenAI Compatibility API

openapi: 3.1.0
info:
  title: LM Studio Local Server Anthropic Compatibility OpenAI Compatibility API
  version: 1.0.0-beta
  description: 'LM Studio runs large language models locally and exposes a local HTTP server with three families of endpoints: the native LM Studio REST API (beta) under `/api/v1`, an OpenAI-compatible surface under `/v1`, and an Anthropic-compatible surface (`/v1/messages`). The server runs on the developer''s own machine (default `http://localhost:1234`). The native `/api/v1` API adds LM Studio-specific capabilities such as MCP via API, stateful chats, model download/load/unload, generation stats (tokens-per-second, time-to-first-token), and optional API-token authentication. Endpoints and paths are transcribed from the public LM Studio developer documentation; request/response bodies follow the documented OpenAI-style message shapes.'
  contact:
    name: LM Studio
    url: https://lmstudio.ai/docs
  license:
    name: Proprietary (LM Studio application)
    url: https://lmstudio.ai/terms
  x-provenance:
    method: generated
    source: https://lmstudio.ai/docs/app/api/endpoints/rest
    generated: '2026-07-20'
servers:
- url: http://localhost:1234
  description: Default local LM Studio server (bind host/port configurable)
tags:
- name: OpenAI Compatibility
  description: OpenAI-compatible endpoints for reuse of existing OpenAI clients.
paths:
  /v1/models:
    get:
      tags:
      - OpenAI Compatibility
      operationId: openaiListModels
      summary: List models (OpenAI-compatible)
      responses:
        '200':
          description: OpenAI-style model list.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ModelList'
  /v1/chat/completions:
    post:
      tags:
      - OpenAI Compatibility
      operationId: openaiChatCompletions
      summary: Create chat completion (OpenAI-compatible)
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatRequest'
      responses:
        '200':
          description: OpenAI-style chat completion.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ChatResponse'
        '400':
          $ref: '#/components/responses/BadRequest'
  /v1/completions:
    post:
      tags:
      - OpenAI Compatibility
      operationId: openaiCompletions
      summary: Create text completion (OpenAI-compatible)
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              properties:
                model:
                  type: string
                prompt:
                  type: string
                stream:
                  type: boolean
              required:
              - model
              - prompt
      responses:
        '200':
          description: OpenAI-style completion.
        '400':
          $ref: '#/components/responses/BadRequest'
  /v1/embeddings:
    post:
      tags:
      - OpenAI Compatibility
      operationId: openaiEmbeddings
      summary: Create embeddings (OpenAI-compatible)
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              properties:
                model:
                  type: string
                input:
                  oneOf:
                  - type: string
                  - type: array
                    items:
                      type: string
              required:
              - model
              - input
      responses:
        '200':
          description: Embedding vectors.
        '400':
          $ref: '#/components/responses/BadRequest'
  /v1/responses:
    post:
      tags:
      - OpenAI Compatibility
      operationId: openaiResponses
      summary: Create a response (OpenAI Responses API compatible)
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              properties:
                model:
                  type: string
                input: {}
              required:
              - model
              - input
      responses:
        '200':
          description: Response object.
components:
  schemas:
    ChatRequest:
      type: object
      properties:
        model:
          type: string
          description: Identifier of a loaded model.
        messages:
          type: array
          items:
            $ref: '#/components/schemas/Message'
        stream:
          type: boolean
          default: false
        temperature:
          type: number
        max_tokens:
          type: integer
        tools:
          type: array
          items:
            type: object
        response_format:
          type: object
          description: Structured-output enforcement via JSON schema.
      required:
      - model
      - messages
    ModelList:
      type: object
      properties:
        object:
          type: string
          example: list
        data:
          type: array
          items:
            $ref: '#/components/schemas/Model'
    Error:
      type: object
      properties:
        error:
          type: object
          properties:
            message:
              type: string
            type:
              type: string
            code:
              type: string
    Model:
      type: object
      properties:
        id:
          type: string
        object:
          type: string
        type:
          type: string
          enum:
          - llm
          - embeddings
          - vlm
        arch:
          type: string
        state:
          type: string
          enum:
          - loaded
          - not-loaded
        max_context_length:
          type: integer
    GenerationStats:
      type: object
      description: LM Studio generation performance metrics.
      properties:
        tokens_per_second:
          type: number
        time_to_first_token:
          type: number
        generation_time:
          type: number
        stop_reason:
          type: string
    ChatResponse:
      type: object
      properties:
        id:
          type: string
        model:
          type: string
        choices:
          type: array
          items:
            type: object
            properties:
              index:
                type: integer
              message:
                $ref: '#/components/schemas/Message'
              finish_reason:
                type: string
        stats:
          $ref: '#/components/schemas/GenerationStats'
        usage:
          type: object
          properties:
            prompt_tokens:
              type: integer
            completion_tokens:
              type: integer
            total_tokens:
              type: integer
    Message:
      type: object
      properties:
        role:
          type: string
          enum:
          - system
          - user
          - assistant
          - tool
        content:
          type: string
      required:
      - role
      - content
  responses:
    BadRequest:
      description: The request was malformed or referenced an unavailable model.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/Error'
  securitySchemes:
    ApiToken:
      type: http
      scheme: bearer
      description: Optional API token. The native `/api/v1` server can be configured to require a bearer token; by default the local server accepts unauthenticated requests from the loopback interface.