Upstash

Best Self-hosted Alternatives to Upstash

A curated collection of the 3 best self hosted alternatives to Upstash.

Managed serverless Redis and Kafka platform providing cloud-hosted key-value and event-streaming services with REST APIs, global replication, and pay-as-you-go billing. Used for low-latency caching, queues, pub/sub and real-time/event-driven applications.

Alternatives List

#1
Valkey

Valkey

Valkey is an open-source, Redis-compatible key/value datastore optimized for caching, queues, and other real-time workloads, with clustering, replication, and extensibility.

Valkey screenshot

Valkey is a high-performance data structure server and distributed key/value datastore designed for low-latency, real-time workloads such as caching and message queues. It is a fork of Redis OSS created before Redis’ license change, and aims to remain permissively licensed and community governed.

Key Features

  • Rich native data structures such as strings, hashes, lists, sets, sorted sets, bitmaps, and HyperLogLogs
  • Replication and high availability options, including Sentinel-style monitoring and failover
  • Cluster mode for horizontal scaling and sharding
  • Extensibility via modules (plugins) to add new commands, data types, and access patterns
  • Built-in Lua scripting for server-side logic and atomic workflows
  • Optional TLS support for encrypted client and inter-node communication

Use Cases

  • Application caching and session storage with low latency
  • Real-time counters, leaderboards, rate limiting, and ephemeral state
  • Queueing and lightweight messaging patterns for background work

Limitations and Considerations

  • Primarily memory-focused workloads; persistence and durability characteristics depend on configuration choices
  • Some advanced features (for example RDMA support) may be experimental and platform-limited

Valkey is a strong fit when you need Redis-style commands and data structures with modern operational options like clustering and encryption. It can serve as a drop-in compatible datastore for many Redis OSS deployments while supporting a broad range of real-time patterns.

24.9kstars
1kforks
#2
ejabberd

ejabberd

ejabberd is an Erlang/OTP-based messaging server providing XMPP chat and presence, MQTT broker capabilities for IoT, and SIP services for real-time communications.

ejabberd screenshot

ejabberd is an open-source, scalable real-time messaging platform built on Erlang/OTP. It provides an XMPP server for chat and presence, and can also act as an MQTT broker and SIP service for broader real-time and IoT workloads.

Key Features

  • XMPP server with support for group chat, publish-subscribe, and many protocol extensions
  • MQTT broker functionality for lightweight device and IoT messaging
  • SIP service support and interoperability options for real-time communications
  • Native clustering designed for high concurrency and large deployments
  • Security-focused design with modern TLS support and configurable authentication options
  • Multiple deployment options including packages and container images

Use Cases

  • Team or consumer chat systems using XMPP (including multi-user chat)
  • IoT telemetry and device messaging using MQTT
  • Real-time communication backends that combine messaging and signaling needs

Limitations and Considerations

  • Full capabilities often depend on correct selection and configuration of modules and protocol extensions
  • Running large clusters typically requires operational expertise in tuning, monitoring, and database/storage choices

ejabberd is a mature, extensible platform suited to organizations needing a reliable messaging core at scale. Its multi-protocol approach makes it useful for both classic chat deployments and modern IoT-oriented messaging architectures.

6.6kstars
1.5kforks
#3
Redict

Redict

Redict is a distributed key/value datastore (a Redis OSS 7.2.4 fork) for caching, queues, and fast data access with clustering, replication, and optional TLS.

Redict screenshot

Redict is a distributed key/value store (a remote dictionary) designed for low-latency data access over the network. It is an independent fork of Redis OSS 7.2.4 and aims to provide a compatible server and CLI for common Redis-style workflows.

Key Features

  • In-memory key/value data model with networked client/server operation
  • Replication and high-availability patterns (including Sentinel-style monitoring)
  • Cluster mode for distributing data across multiple nodes
  • Built-in CLI for interactive administration and troubleshooting
  • Optional TLS support for encrypted connections (OpenSSL-based builds)

Use Cases

  • Application caching and session storage
  • Lightweight message queues and background job coordination
  • Shared state and fast lookup storage for distributed systems

Limitations and Considerations

  • Feature set and module ecosystem compatibility may differ from upstream Redis, depending on divergence from the fork point and supported APIs

Redict is a practical choice for teams that need a Redis-compatible datastore with clustering and optional TLS. It fits well as a foundational building block for performance-sensitive applications and distributed architectures.

Why choose an open source alternative?

  • Data ownership: Keep your data on your own servers
  • No vendor lock-in: Freedom to switch or modify at any time
  • Cost savings: Reduce or eliminate subscription fees
  • Transparency: Audit the code and know exactly what's running