Skip to content

The retrieval control plane

One search API.Four engines.Pick per index.

Headband puts Meilisearch, Typesense, Elasticsearch and Visor behind one REST API, one set of keys and one dashboard. Put each index on the engine that fits it. Give your agents one MCP endpoint that searches them all.

Start freeRead the docs
Eight indexes, one API. Each request goes to the engine its index lives on.About 5 ms added over the raw engine in our benchmark.

The engines

Every retrieval job has a best engine. Run them side by side.

  1. Meilisearch

    For catalogs and your first index.

    Typo tolerance is on by default, and every setting applies to a live index.

  2. Typesense

    For autocomplete and search-as-you-type.

    Set the schema once and it stays out of the way.

  3. Elasticsearch

    For documents, RAG and large corpora.

    Native vectors with kNN, fused with keyword results by Headband, on the free license.

  4. Visor

    For search boxes that take short and long queries.

    ZVN's own engine, run by Headband. One index answers as-you-type in instant mode and long queries with BM25F retrieval, picked per request.

Same API, same filter syntax, same keys. Moving an index is a reindex, not a rewrite. Bring your own engine when an index has to stay in your infrastructure.

Compare the enginesRun one query on all four

For agents

Retrieval for your app and your agents.

One MCP endpoint lets your agents search every index on every engine and get back one ranked list.

Quickstart

From the terminal to your first search.

Drop in the React SDK for the UI, or point your agent at the MCP endpoint. Search keys are browser-safe; admin keys stay on your server.

Getting startedReact SDK
TerminalHeadband CLI
$ npm install -g @headband/cli$ headband login$ headband indexes create products --engine meilisearch$ headband docs push products.json --index products$ headband search products "red wine"
Or call the REST API directly: POST /v1/search

In early access

Engines that sleep when you don't need them.

Early accessDedicated Meilisearch engines park when idle and wake on the next request in about 100 ms. A parked engine uses no compute.

Early accessServerless Typesense runs in a microVM. It parks when idle and wakes fast on the next request.

Early accessVisor runs on dedicated engines, serverless or always on. A serverless Visor engine parks when idle and wakes on the next request.

In developmentElasticsearch in microVMs. In our prototype, an engine woke from a memory snapshot in about 100 ms, and a frozen engine resumed in about 10 ms.

Request early access

Search that follows your database.

Early accessConnect a Postgres database and Headband reads its change stream. You define each document in SQL. Headband re-renders what changed and rebuilds indexes in the background, then swaps them in without downtime.

No sync scripts, no app changes, and Headband never holds your owner credentials.

Request early access

Integrations

Source → sync → route → experiment → ship.

Use Headband with the framework, database, commerce stack and analytics loop you already have. Live SDK support, recipes and roadmap items are labeled separately.

Explore integrations
Frameworks
5recipes
Sync connectors
6patterns
A/B/C tests
3playbooks

Pricing

Simple, predictable pricing.

All four engines on every plan, one of each on Free. Bring your own Meilisearch, Typesense or Elasticsearch on any of them.

Free

$0a month

For side projects and prototypes

  • Every engine, one of each, smallest size
  • 25,000 searches a month
  • 100 MB of stored data
Get started free

Starter

$12a month

For a small production app

  • Every engine, as many as you need
  • 250,000 searches a month
  • 250 MB of stored data
Choose Starter

Pro

$69a month

For growing products

  • Every engine, as many as you need
  • 2M searches a month
  • 2 GB of stored data
Choose Pro

Scale

$249a month

For busy production search

  • Every engine, as many as you need
  • 20M searches a month
  • 10 GB of stored data
Choose Scale
See full pricing details

Start free.

All four engines, no engine to run. Free to start. Paid plans from $12 a month.

Create your account

Need dedicated engines or Postgres sync? Ask for early access.