Momento is enterprise Valkey for in-memory caching: a managed AWS service or a self-hosted Kubernetes stack, with pub/sub messaging and usage-based pricing.
Momento keeps application reads fast as traffic grows by running Valkey caches that a team would otherwise size, patch and fail over by hand. It is the one product of Momento Inc., an independent company, and the same stack serves its cloud platform and its self-hosted software. Existing Valkey and Redis clients connect to it. On the cloud side, Momento Cache comes in two shapes. Flex lets a team set capacity while Momento chooses the topology, and Cluster lets the team pick the instance type, the shards and the replicas. Topics adds serverless pub/sub messaging for servers and devices. Hashes with per-field TTLs, lists, sets and sorted sets live on the server, and a healthy replica takes over when a primary is lost. On the self-hosted side, the Valkey Operator runs sharded clusters on Kubernetes with roles scoped to key patterns. The Valkey Enterprise Image ships patched builds with optional Data Tiering and Observability modules, and the Valkey Router pools connections, coalesces requests and terminates TLS. Momento suits platform and backend engineers whose cache has become critical infrastructure rather than a side component. Companies already on AWS can hand the cluster to Momento's managed service. Teams that keep data inside their own Kubernetes clusters can take the operator, the image and the router instead. Those who need JSON documents, Search indexes or Bloom filters add the modules through an Enterprise support plan, and support levels scale from hobby projects to mission-critical workloads. In practice Momento sits in front of a primary database, answering repeated reads from memory so the database carries less load. Game studios use Topics for in-game chat and sorted sets for leaderboards, while media teams push live updates to viewers during streamed events. Operators running the self-hosted line deploy it through Kubernetes resources and feed its metrics into an OpenTelemetry pipeline they already watch. The result is one Valkey stack a team can rent or run, depending on where its data has to live.
Pricing
View plans-
Momento Cache Flex
$13 per month
Usage-based: from $13 per GB-month of configured capacity
- Flex Standard $0.018 and Flex Performance $0.027 per GiB-hour in us-east-1; other regions may vary
- Billed on maxmemory summed over primaries and replicas
- 3 shards with 2 replicas each and 3 GiB per node on Performance comes to $174.96 for 30 days
- Fully managed Valkey; you set capacity in GiB and Momento picks the topology
- 200 GB of data transfer a month included, then $0.05 per GB
- Limited preview, access on request; runs on AWS
-
Momento Cache Cluster
Custom pricing
- From $0.0105 per instance-hour (t4g.micro), primaries and replicas counted
- Other instance types at their AWS list price times 1.25
- One t4g.micro node for 30 days comes to $7.56
- You choose instance type, shard count and replicas per shard
- 200 GB of data transfer a month included, then $0.05 per GB; no included transfer for multi-account organizations or enterprise-tier accounts
- Limited preview, access on request; runs on AWS
-
Topics
Custom pricing
- From $1 per 1M operations
- Serverless pub/sub messaging for any number of servers and devices
-
Valkey Operator (self-hosted)
$3.50 per month
Usage-based: $3.50 per GB-month of managed capacity
- Kubernetes operator for Valkey clusters with CRDs, RBAC policy and rolling updates
- Runs in your own cluster; in-memory only, with no backup or restore and asynchronous replication
Features
- Included: Rich data structures beyond strings
- Included: Per-key expiration and eviction policies
- Not included: Snapshot persistence to disk
- Not included: Append-only log persistence
- Included: Replica failover with automatic promotion
- Included: Horizontal clustering of the keyspace
- Not included: Active-active geo replication
- Included: Pub/sub messaging channels
- Not included: Atomic server-side scripting
- Included: SSD tiering beyond available RAM
- Included: Secondary indexes on stored values
- Included: Native JSON value type
- Not included: Multi-threaded request handling
- Included: Per-user command and key ACLs
- Not included: Client-side caching invalidation
- Not included: Streams with consumer groups
- Not included: Keyspace memory reporting
Additional Features
- Momento Cache Flex with a Momento-picked topology
- Momento Cache Cluster with chosen instance, shards and replicas
- Momento Topics serverless messaging
- Valkey Operator for Kubernetes
- Valkey Enterprise Image with CVE-patched builds
- Data Tiering Module to NVMe, S3 Express or Postgres
- Observability Module with OpenTelemetry metrics
- Valkey Router connection pooling and request coalescing
Best for
- Platform teams that want dedicated Valkey capacity without operating it
- Teams running Valkey on their own Kubernetes clusters
- Applications already hosted on AWS
- Game studios running chat and leaderboards
- Media and live streaming services with live-event traffic
- Backend teams that need JSON or search modules under an enterprise support plan
Use cases
- Caching reads in front of a primary database
- Fanning out in-game chat messages
- Pushing real-time updates during live streams
- Ranking players on leaderboards with sorted sets
- Running sharded Valkey clusters on Kubernetes
- Pooling client connections in front of Valkey
- Spilling cold string data to NVMe or S3 Express
- Indexing hashes and JSON for filtered queries
- Restricting service accounts to key patterns with ACLs
- Patching Valkey images against CVEs
Screenshots & Videos
Explore Momento in action
User Reviews
Compare with Similar Software
Click to compare side by side with Momento
