Roboflow Model API

openapi: 3.2.0
info:
  title: Roboflow Inference Server Model API
  description: Roboflow inference server
  termsOfService: https://roboflow.com/terms
  contact:
    name: Roboflow Inc.
    url: https://roboflow.com/contact
    email: help@roboflow.com
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0.html
  version: 1.3.8
tags:
- name: Model
paths:
  /model/registry:
    get:
      summary: Get model keys
      description: Get the ID of each loaded model
      operationId: registry_model_registry_get
      responses:
        '200':
          description: Successful Response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ModelsDescriptions'
      tags:
      - Model
components:
  schemas:
    ModelsDescriptions:
      properties:
        models:
          items:
            $ref: '#/components/schemas/ModelDescriptionEntity'
          type: array
          title: Models
          description: List of models that are loaded by model manager.
        total_vram_bytes:
          anyOf:
          - type: integer
          - type: 'null'
          title: Total Vram Bytes
          description: Total estimated VRAM consumed by all loaded models in bytes.
        gpu_memory_used:
          anyOf:
          - type: integer
          - type: 'null'
          title: Gpu Memory Used
          description: Current GPU memory in use in bytes (device-level, includes all runtimes).
        gpu_memory_total:
          anyOf:
          - type: integer
          - type: 'null'
          title: Gpu Memory Total
          description: Total GPU memory available in bytes.
        torch_cuda_allocated:
          anyOf:
          - type: integer
          - type: 'null'
          title: Torch Cuda Allocated
          description: Live tensor memory allocated by PyTorch's CUDA allocator in bytes.
        torch_cuda_reserved:
          anyOf:
          - type: integer
          - type: 'null'
          title: Torch Cuda Reserved
          description: Total memory reserved by PyTorch's CUDA allocator in bytes.
        torch_cuda_allocator_cache:
          anyOf:
          - type: integer
          - type: 'null'
          title: Torch Cuda Allocator Cache
          description: Reserved but currently unallocated PyTorch CUDA memory in bytes.
        non_torch_gpu_memory:
          anyOf:
          - type: integer
          - type: 'null'
          title: Non Torch Gpu Memory
          description: Device memory not reserved by PyTorch in bytes. This includes native runtimes,
            CUDA context overhead, and allocations from other processes.
      type: object
      required:
      - models
      title: ModelsDescriptions
    ModelDescriptionEntity:
      properties:
        model_id:
          type: string
          title: Model Id
          description: Identifier of the model
          examples:
          - some-project/3
        task_type:
          type: string
          title: Task Type
          description: Type of the task that the model performs
          examples:
          - classification
        batch_size:
          anyOf:
          - type: integer
          - type: 'null'
          title: Batch Size
          description: Batch size accepted by the model (if registered).
        input_height:
          anyOf:
          - type: integer
          - type: 'null'
          title: Input Height
          description: Image input height accepted by the model (if registered).
        input_width:
          anyOf:
          - type: integer
          - type: 'null'
          title: Input Width
          description: Image input width accepted by the model (if registered).
        vram_bytes:
          anyOf:
          - type: integer
          - type: 'null'
          title: Vram Bytes
          description: Estimated GPU VRAM consumed by this model in bytes (measured during load).
        request_aliases:
          items:
            type: string
          type: array
          title: Request Aliases
          description: Other model IDs that resolved to this model.
        request_paths:
          items:
            type: string
          type: array
          title: Request Paths
          description: HTTP request paths that triggered inference on this model (e.g. /door-glyph-locator/10,
            /infer/object_detection).
      type: object
      required:
      - model_id
      - task_type
      title: ModelDescriptionEntity