Superlinked

Superlinked is an open-source Python framework for building vector-compute pipelines that encode structured and unstructured data into multi-modal embeddings for retrieval, recommendations, RAG, and analytics. When deployed via the Superlinked Server, the framework auto-generates a REST API - ingestion and query endpoints derived directly from your schema, index, and query definitions - and connects to external vector databases. A managed Superlinked Cloud is available in early access.

5 APIs 0 Features

VectorsEmbeddingsVector SearchRetrievalRecommendationsRAG

APIs

Superlinked Ingestion API

Schema-generated REST ingestion endpoint exposed by the Superlinked Server. Each schema defined in your app produces a POST /api/v1/ingest/{schema} endpoint that accepts records...

Superlinked Query API

Schema-generated REST query endpoint exposed by the Superlinked Server. Each RestQuery you register produces a POST /api/v1/search/{query_name} endpoint whose request body param...

Superlinked Data Loader API

Server endpoints for triggering and monitoring configured batch data loaders - POST /data-loader/{name}/run starts a load and GET /data-loader/{name}/status reports progress - u...

Superlinked Framework (Python)

The open-source Apache-2.0 Python framework (pip install superlinked) used to declare Schema, Space, Index, Query, Source, and Executor objects. The same definitions run in-memo...

Superlinked Cloud

Managed, production-scale hosting of Superlinked-powered apps, available in early access via a sales-led demo. Runs the same schema-generated REST surface as the self-hosted Sup...

Collections

Superlinked Server API

OPEN

Pricing Plans

Superlinked Plans Pricing

2 plans

PLANS

Rate Limits

Superlinked Rate Limits

4 limits

RATE LIMITS

FinOps

Superlinked Finops

FINOPS

Resources

🔗

VulnerabilityDisclosure

VulnerabilityDisclosure

Sources

opencollection: 1.0.0
info:
  name: Superlinked Server API
  version: '1.0'
request:
  auth:
    type: apikey
    apikey:
      key: x-api-key
      value: '{{apiKey}}'
      in: header
items:
- info:
    name: Ingestion
    type: folder
  items:
  - info:
      name: Ingest records for a schema
      type: http
    http:
      method: POST
      url: '{{baseUrl}}/api/v1/ingest/:schema'
      params:
      - name: schema
        value: paragraph
        type: path
        description: Identifier of the Superlinked schema whose ingest endpoint is being called.
      body:
        type: json
        data: "{\n  \"id\": \"doc-1\",\n  \"body\": \"Example text to embed.\",\n  \"rating\": 5\n}"
    docs: Schema-generated endpoint. Accepts records matching the schema; the server computes embeddings and writes them to
      the connected vector database.
- info:
    name: Query
    type: folder
  items:
  - info:
      name: Run a registered query
      type: http
    http:
      method: POST
      url: '{{baseUrl}}/api/v1/search/:query_name'
      params:
      - name: query_name
        value: paragraph_query
        type: path
        description: Name of the registered RestQuery to execute.
      body:
        type: json
        data: "{\n  \"natural_query\": \"comfortable hotel near the beach\",\n  \"limit\": 10\n}"
    docs: Schema-generated endpoint. Request body parameters are derived from the registered query definition. Returns ranked
      results from the connected vector store.
- info:
    name: Data Loader
    type: folder
  items:
  - info:
      name: Trigger a configured data loader
      type: http
    http:
      method: POST
      url: '{{baseUrl}}/data-loader/:name/run'
      params:
      - name: name
        value: review
        type: path
        description: Name of the configured data loader.
    docs: Starts a configured batch data loader to backfill vectors from a file or external source.
  - info:
      name: Get data loader status
      type: http
    http:
      method: GET
      url: '{{baseUrl}}/data-loader/:name/status'
      params:
      - name: name
        value: review
        type: path
        description: Name of the configured data loader.
    docs: Reports the progress and state of a configured data loader.
bundled: true