Services: Custom Search API & Search Engines

Build your own domain-specific search engine with the same tech that powers the 4search network. Go far beyond the capablilites of other custom seearch engine queries with advanced taxonomies and ontologies, combine classic text ranking with dense vector embeddings, and ship production-grade relevance fast.

Request Access See Capabilities

Why 4search Search Services?

  • Hybrid Retrieval: BM25 + dense vector (KNN) + rerankers for SOTA relevance.
  • Taxonomies & Ontologies: Query by categories, entities, and relationships (parents/children/siblings).
  • Structured + Unstructured: Filter by metadata (dates, sources, regions) and free text simultaneously.
  • Production Ready: Modern and simple APIs. High volume.

Under the hood, we support Elasticsearch/OpenSearch-style indices, dense vectors (8-bit or float), and pluggable rerankers. Bring your own models, or use our defaults.

Ideal For

  • Vertical search portals (finance, bio, sports, local, etc.)
  • Enterprise knowledge search & intranet
  • E-commerce discovery & faceting
  • Media archives & editorial research
  • AI copilots & retrieval-augmented generation (RAG)

Core Capabilities

Advanced Querying

  • Hybrid text + vector search
  • Facet filters & range filters
  • Entity/knowledge-graph expansion
  • Synonyms, spell-correction, boosts/decay

Taxonomy & Ontology

  • Hierarchical categories (multi-level)
  • Ontological relations (is-a, part-of, related-to)
  • Per-node boosts, exclusions, and routing
  • Auto-classification pipelines available

Embeddings & Rerankers

  • 768-D (or custom) dense vectors
  • Quantized (8-bit) or float32
  • Cross-encoder or LLM rerankers
  • RAG-friendly result payloads

Indexing & Feeds

  • Bulk loaders (JSON/NDJSON/CSV)
  • Scheduled crawls & webhooks
  • De-duplication & canonicalization
  • Field-level mapping templates

Relevance & Analytics

  • Click/engagement signals
  • A/B experiments & eval sets
  • Curations, pins, and hard rules
  • Telemetry & dashboards

APIs & UI Widgets

  • REST/JSON & GraphQL endpoints
  • Autocomplete & suggestions
  • Drop-in search box & results UI
  • SSO & rate-limit controls

Integration Options

Hosted by 4search

  • Managed infrastructure & scaling
  • SLAs, monitoring, and support
  • Role-based access control & SSO
  • Versioned configs and rollbacks

Self-Hosted / Hybrid

  • Deploy on your cloud or on-prem
  • We deliver schemas, pipelines, and tooling
  • Optional managed embeddings service
  • Periodic sync with 4search indices (optional)

Front-end widgets (search box, facets, result cards) are available in vanilla JS and React variants and can be themed to match your brand. We also provide example templates for Bootstrap 5.

What We Deliver

Search Stack

  • Index schemas & mappings (text + vectors)
  • Ingestion pipelines & dedupe rules
  • Query DSL presets (hybrid, faceting, etc.)
  • Reranker & synonyms configuration

Docs & Support

  • API & SDK references
  • Relevance playbooks & eval sets
  • Launch checklist & observability hooks
  • Training for your team (optional)

Pricing

Engagement Models

  • Fixed-scope implementation
  • Monthly subscription for hosted search
  • Usage-based API (queries, storage, bandwidth)
  • Enterprise support & SLA tiers

Options & Add-Ons

  • Embeddings hosting & model upgrades
  • Custom taxonomy/ontology design
  • RAG adapters for popular LLM stacks
  • Data licensing from the 4search corpus

For a tailored quote, share your corpus size, expected QPS, geography, and security requirements.

Terms

  • License & IP: You retain rights to your content; configurations and code delivered under your project agreement.
  • Data Options: Bring-your-own corpus, license 4search data, or blend both (subject to data licensing terms).
  • Security: SSO/SAML/OIDC supported; VPC peering and private networking available.
  • Compliance: Best-effort alignment with robots and publisher guidelines for web content; enterprise data stays in your tenancy.
  • Support: Email support included; premium SLA options for hosted deployments.

Ready to build your custom search engine?

Tell us about your corpus, users, and KPIs. We’ll propose the best architecture and plan.

Contact Sales View Pricing