Search for your app and your AI agents.
A hosted search API with a client-side mode and an MCP server. Unlimited searches on every plan: you pay for what you index, never per query.
Sprigr Search is a separate developer product. Looking for the operations hub for your business? See the Sprigr platform.
One engine, three ways in.
REST for your app, the browser for instant keyword search, MCP for agents. The same indexes behind all three.
-
Hosted REST API
Create indexes, import JSON objects and query them over HTTPS. Typo tolerance, filters, facets and aggregations are built in.
-
Client-side mode
Download an index once and a WebAssembly engine answers keyword queries in the browser, with no request per search.
-
Hybrid semantic search
Turn on
semantic_searchfor an index and Sprigr adds vector matching to keyword ranking. No embedding model to run. -
MCP server for AI agents
Add
@sprigr/mcp-serverto an MCP client and an agent can create indexes, store records and search them. -
Scoped keys
Limit keys to named indexes, or mint secured keys with a locked filter so each end user only sees their own rows.
-
Change subscriptions
Subscribe a webhook to an index and Sprigr posts an event to it when the index changes, with delivery health you can check.
What developers build with it.
-
MCP search server
Searchable cloud storage for AI agents, connected over the Model Context Protocol.
-
Claude MCP search
Persistent, searchable memory for Claude Desktop and Claude Code in one config block.
-
AI agent storage
Store JSON records, search them and fetch them by ID over REST or MCP.
-
Client-side search
A WebAssembly engine that answers keyword queries in the browser.
-
Next.js search
Instant search in a Next.js app without API routes or a search backend.
-
Static site search
Search for Hugo, Jekyll, Eleventy, Astro and any static site generator.
-
Search without API calls
Users search locally, so there is no per-query cost at any traffic level.
-
Privacy-first search
Keyword queries stay in the browser and are never sent to a server.
Against other search engines.
Honest head to heads with the hosted and self-hosted engines you are likely weighing up.
-
Sprigr vs Algolia
Flat per-object pricing against per-search billing.
-
Sprigr vs Typesense
Hosted with a client-side mode against a self-hosted engine.
-
Sprigr vs Meilisearch
No infrastructure to run against open source plus paid cloud.
-
Elasticsearch alternative
Site and app search without a JVM cluster to operate.
-
Stork Search alternative
Client-side search with hosting, analytics and an admin dashboard.
-
All search engines
Algolia, Typesense, Meilisearch and Sprigr side by side.
Index your first records in minutes.
The Free plan has no expiry. Upgrade when your index outgrows it.