const cache = new Keyv() await cache.set('greeting', 'hello', 1000) const value = await cache.get('greeting')


Use a Redis adapter:

```bash
npm install @keyvhq/core @keyvhq/redis

const Keyv = require('@keyvhq/core')
const KeyvRedis = require('@keyvhq/redis')

const cache = new Keyv({
  store: new KeyvRedis('redis://user:pass@localhost:6379'),
  namespace: 'app-cache'
})

Recommended Workflow

Start with @keyvhq/core in memory for local development.
Add a storage adapter only when persistence or shared cache is needed.
Set namespace per module to avoid collisions and accidental global clears.
Use TTL in milliseconds, either globally (ttl option) or per set.
Keep cache operations behind one service/module so adapter changes stay isolated.

Core API

new Keyv(options): create instance.
set(key, value, ttl?): store value, optional TTL in milliseconds.
get(key): read value.
has(key): check existence.
delete(key): remove one key.
clear(): remove all keys in the current namespace.
iterator(): async iterate entries (avoid for large datasets).

Common Options

store: adapter instance (default is in-memory Map).
namespace: key namespace to isolate data.
ttl: default TTL in milliseconds.
serialize / deserialize: custom serialization for advanced types.
raw: return internal stored object including expiry metadata.

Adapter Selection

Choose an official adapter based on runtime and infrastructure:

@keyvhq/redis: low-latency shared cache, easy horizontal scaling.
@keyvhq/mongo: Mongo-backed cache persistence.
@keyvhq/mysql: MySQL or MariaDB-backed storage.
@keyvhq/postgres: PostgreSQL-backed storage.
@keyvhq/sqlite: file-based local persistence.
@keyvhq/file: lightweight JSON/file storage.
keyv-s3: S3 object storage adapter for large, low-cost cache persistence.

Connector Setup Snippets

Use one connector at a time as store:

Redis (`@keyvhq/redis`)

npm install @keyvhq/core @keyvhq/redis

const Keyv = require('@keyvhq/core')
const KeyvRedis = require('@keyvhq/redis')

const cache = new Keyv({
  store: new KeyvRedis('redis://user:pass@localhost:6379'),
  namespace: 'app-cache'
})

Mongo (`@keyvhq/mongo`)

npm install @keyvhq/core @keyvhq/mongo

const Keyv = require('@keyvhq/core')
const KeyvMongo = require('@keyvhq/mongo')

const cache = new Keyv({
  store: new KeyvMongo('mongodb://user:pass@localhost:27017/dbname'),
  namespace: 'app-cache'
})

MySQL (`@keyvhq/mysql`)

npm install @keyvhq/core @keyvhq/mysql

const Keyv = require('@keyvhq/core')
const KeyvMySQL = require('@keyvhq/mysql')

const cache = new Keyv({
  store: new KeyvMySQL('mysql://user:pass@localhost:3306/dbname'),
  namespace: 'app-cache'
})

PostgreSQL (`@keyvhq/postgres`)

npm install @keyvhq/core @keyvhq/postgres

const Keyv = require('@keyvhq/core')
const KeyvPostgres = require('@keyvhq/postgres')

const cache = new Keyv({
  store: new KeyvPostgres('postgresql://user:pass@localhost:5432/dbname'),
  namespace: 'app-cache'
})

SQLite (`@keyvhq/sqlite`)

npm install @keyvhq/core @keyvhq/sqlite

const Keyv = require('@keyvhq/core')
const KeyvSQLite = require('@keyvhq/sqlite')

const cache = new Keyv({
  store: new KeyvSQLite('sqlite://path/to/database.sqlite'),
  namespace: 'app-cache'
})

File (`@keyvhq/file`)

npm install @keyvhq/core @keyvhq/file

const Keyv = require('@keyvhq/core')
const KeyvFile = require('@keyvhq/file')

const cache = new Keyv({
  store: new KeyvFile({ filename: './.cache/keyv.json' }),
  namespace: 'app-cache'
})

S3 (`keyv-s3`)

npm install @keyvhq/core keyv-s3 @aws-sdk/client-s3

const Keyv = require('@keyvhq/core')
const KeyvS3 = require('keyv-s3')

const cache = new Keyv({
  store: new KeyvS3({
    region: 'us-east-1',
    namespace: 'app-cache',
    accessKeyId: process.env.S3_ACCESS_KEY_ID,
    secretAccessKey: process.env.S3_SECRET_ACCESS_KEY
  })
})

Decorators

Use decorators for specialized behavior:

@keyvhq/compress: compress stored payloads.
@keyvhq/memoize: memoize function calls through Keyv.
@keyvhq/multi: combine local and remote stores.
@keyvhq/offline: add offline-aware behavior.
@keyvhq/stats: collect usage metrics over time.

Integration Pattern For Libraries

When adding cache support to a module:

Expose a cache option in module configuration.
Accept any Keyv-compatible store.
Default to in-memory behavior when no cache is passed.
Set a dedicated namespace before calling clear().

function createClient({ cache = new Keyv({ namespace: 'my-module' }) } = {}) {
  return {
    async getOrFetch(key, fetcher, ttl) {
      const cached = await cache.get(key)
      if (cached !== undefined) return cached

      const value = await fetcher()
      await cache.set(key, value, ttl)
      return value
    }
  }
}

Reliability Notes

TTL values are milliseconds.
clear() without a namespace can wipe all entries in that store.
Storage adapter errors should be handled close to cache boundaries.
For bulk iteration or analytics, prefer native database tooling over iterator().

References

Keyv monorepo: https://github.com/microlinkhq/keyvhq
Root docs: https://github.com/microlinkhq/keyvhq/blob/master/README.md
Keyv S3 adapter: /Users/kikobeats/Projects/microlink/keyv-s3

Recommended Workflow

Core API

Common Options

Adapter Selection

Connector Setup Snippets

Redis (@keyvhq/redis)

Mongo (@keyvhq/mongo)

MySQL (@keyvhq/mysql)

PostgreSQL (@keyvhq/postgres)

SQLite (@keyvhq/sqlite)

File (@keyvhq/file)

S3 (keyv-s3)

keyvhq

keyvhq

Quick Start