specification: API Commons Vocabulary
specificationVersion: '0.1'
provider: Vespa
providerId: vespa-ai
created: '2026-05-25'
modified: '2026-05-25'
description: Vespa domain vocabulary mapping operational dimensions, capabilities, and terminology.
tags:
  - AI Search
  - Vector Database
  - Big Data Serving
concepts:
  - name: Application Package
    description: A versioned bundle of Vespa schemas, services.xml, search definitions, rank
      profiles, and components deployed to a Vespa application.
  - name: Schema
    description: A Vespa document type definition (fields, indexes, summary classes) declared in
      `.sd` files within an application package.
  - name: Document
    description: A unit of data stored in a Vespa content cluster, identified by namespace, document
      type, and document id.
  - name: Namespace
    description: A logical multi-tenant segment within a Vespa document type. Used to isolate
      tenants or sub-domains within a single application.
  - name: Content Cluster
    description: A cluster of Vespa storage and search nodes that index, persist, and query
      documents.
  - name: Container Cluster
    description: A cluster of Vespa container nodes that host the Query API, Document API, custom
      searchers, and document processors.
  - name: YQL
    description: Vespa Query Language — Vespa's SQL-like query language used to express selection,
      ranking, and grouping for searches.
  - name: Rank Profile
    description: A named ranking configuration in a schema that defines first-phase, second-phase,
      global-phase, and match-phase expressions and ML model inputs.
  - name: Tensor
    description: Vespa's native multi-dimensional numeric data type used for vectors, matrices, and
      ML model inputs/outputs at serving time.
  - name: HNSW
    description: Hierarchical Navigable Small World — Vespa's approximate-nearest-neighbor index
      for vector search.
  - name: Nearest Neighbor
    description: A Vespa query operator that finds documents whose tensor field is closest to a
      query tensor by a configured distance metric.
  - name: Streaming Search
    description: A Vespa search mode that scans a user's documents on demand instead of using a
      shared inverted index. Ideal for personal-data search.
  - name: Visit
    description: Bulk document iteration over a content cluster using the `/document/v1/` API with
      continuation tokens.
  - name: Selection Expression
    description: A Vespa expression language used to filter documents in visit, copy, update-where,
      and delete-where operations.
  - name: Application Endpoint
    description: A public endpoint for a Vespa Cloud application, authenticated by mTLS or data-
      plane token.
  - name: Zone
    description: A Vespa Cloud deployment region/environment (e.g. dev, perf, prod) within a cloud
      provider.
services:
  - Vespa Query API
  - Vespa Document API
  - Vespa Deploy API
  - Vespa Tenant API
  - Vespa Config API
  - Vespa Cluster Controller API
  - Vespa State API
  - Vespa Metrics API
tools:
  - Vespa CLI (Go)
  - pyvespa (Python)
  - vespa-feed-client (Java)
  - Vespa Docker Image
  - Vespa Prometheus Exporter
standards:
  - Apache 2.0 License
  - OpenAPI 3.0
  - Prometheus exposition format
  - JSON Lines
  - ONNX
  - mTLS