Cuely is now open source. See exactly what it receivesCode on GitHub

Facts read November 20, 2026

Algolia vs Postgres full-text search

Postgres full-text search is Cuely's default search pick when your description doesn't favor either one: no new vendor or sync job: search the rows where they live, with tenant filters as a plain WHERE clause. Pick Algolia if you care most about high volume.

Side by side

Fact
What it isA hosted search API with instant, typo-tolerant results and ready-made UI libraries.Search built into Postgres (tsvector, tsquery and GIN indexes), in the database you already run.
Free planFree: 10K search requests/month, 50K records, 5K recommendation requests/month. No credit card required.Free: part of Postgres; you pay only for the database you already have.
Paid fromGrow: 10K search requests/month included then $0.50 per additional 1K; 100K records included then $0.40 per additional 1K. Billed monthly, no contract.Not stated
At 1,000 users$20/month$0/month
Open sourceNoYes
Self-hostableNoYes
Best forInstant search as people type; Storefronts and catalogs with facets; Teams that don't want to run a search serverA search bar over tens of thousands of rows; Filtering search results by tenant or permissions in SQL; Fewest moving parts
Watch outBilled per search request: search-as-you-type sends one request per keystroke unless you debounce. Free stops at 10K search requests a month. Records must be synced from your database to Algolia.No typo tolerance out of the box; the pg_trgm extension adds fuzzy matching. Relevance tuning and facet counts take hand-written SQL. Search load runs on your database, so heavy search traffic competes with everything else.

Where they differ

The published scores, from −3 to +3, that decide which one Cuely picks for each need.

NeedAlgoliaPostgres full-text search
prefers open source or self-hosting−2 Closed source and hosted only.+3 Part of open-source Postgres, on any host.
keep costs as low as possible−1 Per-search billing: 50,000 searches a month is about $20, and search-as-you-type multiplies requests.+3 No separate bill; it uses the database you already pay for.
fewer vendors0 No particular strength here.+3 Search stays in your existing Postgres; no new vendor or index to keep in sync.
high volume+1 Fully managed: large indexes and traffic need no servers to size.−1 Past millions of rows or heavy search traffic, a dedicated search engine is easier to tune and keeps load off the database.

Sources: Algolia pricing (read November 20, 2026), Postgres full-text search pricing (read November 20, 2026). No vendor pays for placement.

Other search options