v5

latestOpenAPI 3.1.02026-08-025631,1012.8 MB

Discovery

List Available Feature Extractors

Discover all available feature extractors with their capabilities, supported modalities, output features, and example usage. Use this to understand what extractors are available when configuring namespaces and collections in manifests.

get/v1/discovery/extractors

Response

Successful Response

namestring required

Feature extractor name (e.g., 'multimodal_extractor')

versionstring required

Feature extractor version (e.g., 'v1')

descriptionstring required

Human-readable description of what this extractor does

supported_modalitiesstring[]

List of supported input modalities (text, image, video, audio)

output_featuresobject[]

List of features produced by this extractor

input_schemaobject nullable

JSON Schema for extractor inputs — what fields the extractor reads from source objects

parameter_schemaobject nullable

JSON Schema for tunable parameters (defaults, ranges, descriptions for every knob)

output_schemaobject nullable

JSON Schema for output documents — what fields appear in extracted documents

type_modestring nullable

What input types this extractor can handle: 'type_specific' (only one type, e.g. video-only) or 'multimodal' (handles multiple types with conditional processing). Type-specific extractors cannot use automatic-typed bucket properties.

expected_input_typesobject nullable

For type-specific extractors: maps input keys to required types (e.g., {'video': 'video', 'thumbnail': 'image'}). For multimodal extractors: null.

inference_typestring nullable

Kind of real-time inference this extractor provides: 'embedding', 'rerank', 'classify', 'generate', or 'general'. Determines which retriever stages are compatible. Null if the extractor is batch-only.

supported_input_typesstring[]

Accepted input types (e.g., ['video', 'image'])

max_inputsobject

Maximum number of inputs per type (e.g., {'video': 1})

default_parametersobject

Default parameter values — use as a starting point for tuning

required_vector_indexesobject[] nullable

Vector indexes produced by this extractor (name, dimensions, distance metric, feature_uri)

position_fieldsstring[]

Fields that uniquely identify each output document within a source object

capabilitiesstring[]

What this extractor can do: 'batch' (feature extraction during ingestion), 'realtime' (query-time inference for retriever stages)

example_usageobject

Minimal working configuration for namespace + collection + input_mappings + parameters

Example response

[
  {
    "costs": {
      "rates": [
        {
          "credits_per_unit": 10,
          "description": "Per minute of video processed",
          "unit": "minute"
        }
      ],
      "tier": 2,
      "tier_label": "STANDARD"
    },
    "description": "Extracts embeddings from text, images, and video frames",
    "example_usage": {
      "collection": {
        "feature_extractor": {
          "input_mappings": {
            "content": "video"
          },
          "name": "multimodal_extractor",
          "parameters": {
            "split_method": "time",
            "time_split_interval": 10
          },
          "version": "v1"
        }
      },
      "namespace": {
        "feature_extractors": [
          {
            "name": "multimodal_extractor",
            "version": "v1"
          }
        ]
      }
    },
    "name": "multimodal_extractor",
    "output_features": [
      {
        "name": "multimodal_embedding",
        "type": "embedding",
        "uri": "mixpeek://multimodal_extractor@v1/multimodal_embedding"
      }
    ],
    "parameter_schema": {
      "properties": {
        "split_method": {
          "default": "time",
          "description": "Video splitting strategy",
          "type": "string"
        },
        "time_split_interval": {
          "default": 10,
          "description": "Interval in seconds for time splitting",
          "type": "integer"
        }
      },
      "type": "object"
    },
    "supported_input_types": [
      "video",
      "image"
    ],
    "supported_modalities": [
      "video",
      "image"
    ],
    "version": "v1"
  }
]