SuggestAPIDocumentation
Contact usOpen app
Getting Started

Overview

Understand how SuggestAPI handles indexing, autocomplete, typeahead, analytics, and operational controls.

Copy page
Open in ChatGPTOpen with page contextOpen in ClaudeOpen with page context

SuggestAPI is a search platform for teams that need fast autocomplete, typeahead, document indexing, and agent-readable discovery in one stack.

The current product surface includes:

  • Hybrid search that combines lexical FTS with semantic vector retrieval
  • Fast precomputed typeahead suggestions
  • Index creation, schema management, merchandising rules, and blue-green promotion
  • Document ingestion by JSON, CSV, JSONL, Google Shopping feeds, and scheduled integrations
  • Personalization, event tracking, and graph-based recommendations
  • An agent gateway for domain knowledge surfaces, agent workflows, and capability discovery

Core workflow

Most teams use SuggestAPI in a loop:

  1. Create an index.
  2. Define schema and ranking behavior.
  3. Ingest documents or connect a sync source.
  4. Query /v1/autocomplete or /v1/typeahead.
  5. Track clicks, conversions, and personalization events.
  6. Tune ranking, merchandising, and integrations over time.

Main building blocks

  • /v1/autocomplete for full ranked autocomplete
  • /v1/typeahead for low-latency suggestion lookups
  • optional filters, personalization, spell correction, synonyms, and vector search

Data management

  • logical indexes with active physical index promotion
  • inline document ingestion and asynchronous bulk ingestion jobs
  • schema, merchandising, export, reindex, and task endpoints

Integrations

  • Google Shopping feed import
  • webhook ingestion paths for ecommerce and CRM flows

Controls and reporting

  • customer usage summaries and daily usage endpoints
  • API key management with public/private modes
  • billing checkout and portal routes
  • plan, file-size, document-limit, and overage controls

Agent Discovery

  • agent.suggestapi.com/oks/{domain} provides a domain-scoped knowledge surface
  • /v1/agent/* supports agent search, comparison, recommendations, and safe commerce handoffs
  • ARD catalogs let agents discover published capabilities without treating discovery as inventory search

For a production-focused view of those constraints, see Service Limits.