Dragonfly

Dragonfly Cloud

Better performance.Lower cost.Fully managed.

Dragonfly is a modern, multi-threaded, in-memory data store built for real-time workloads at scale. Dragonfly Cloud is the fully managed service: pick your cloud and how much memory to provision, and we run the rest. Teams see 25X the performance and reduce costs by 80% versus legacy in-memory data stores.

PricingTalk to us

What you get with Dragonfly Cloud

All the engine. None of the operations.

Dragonfly Cloud pairs the same multi-threaded engine with a fully managed service that scales both vertically and horizontally, so your team ships instead of operating infrastructure.

Performance

Tuned by the engineers who built Dragonfly for your specific workload, so you get maximum throughput without becoming a tuning expert yourself.

Effortless scaling

Same engine, no ceiling. Swarm shards for you across hundreds of TB and hundreds of millions of RPS, with autoscaling as demand moves, so you manage memory capacity, not instances and clusters.

Deploy anywhere

Fully managed with bring-your-own-cloud on AWS, GCP, Azure, or on-prem. Run it in the region of your choice, with data kept inside your own account.

Expert support

24×7 support direct from the engineers who built Dragonfly, backed by contractual SLAs, rather than DIY support from community forums.

Dev tools & monitoring

A management console built in: monitoring, alerts, logs, Infrastructure-as-Code (IaC), and automatic failover with multi-region backups. Or bring your own tools.

Security

RBAC, ACL, and admin audit logs, with data kept inside your own cloud account, ready for security reviews.

Success story

50% faster. 70% fewer nodes.

Instacart migrated its multi-hundred-node ad-serving feature store to Dragonfly Cloud, cutting cluster size roughly 70% (from hundreds of nodes to about 100) and slashing both average and P99 latency by 50%, thanks to Dragonfly's multi-threaded, vertically-scalable architecture.

Read the case study
Dragonfly × Instacart customer story: 50% faster, 70% fewer nodes. How Instacart rebuilt its real-time ad-serving feature store on Dragonfly Cloud.

A closer look

25X more throughput. 80% lower cost. Zero ops.

From performance tuning to security controls, here's exactly what changes when Dragonfly runs as a fully managed service.

Dragonfly vs. Valkey OSS mixed workload benchmark on Google Cloud

Performance

Optimized tuning

No more guessing at CPU, IRQ, network, and memory settings on bare metal. Dragonfly Cloud ships pre-configured and optimized end-to-end for maximum performance, so it runs production-ready from day one instead of depending on your setup.

See the benchmark
Dragonfly Swarm scaling across hundreds of TB with no ceiling

Effortless scaling

Same engine, no ceiling

Manage memory capacity, not instances or clusters. Swarm shards Dragonfly Cloud for you across hundreds of TB and hundreds of millions of RPS, with autoscaling as demand moves, so you deliver a lightning-fast experience on your highest-traffic day on record.

How it scales
Deploy Dragonfly Cloud anywhere: AWS, GCP, Azure, or on-prem

Deploy anywhere

Run on your infrastructure of choice

Deploy Dragonfly Cloud on AWS, GCP, Azure, or on-prem in the region you choose, for a secure, low-latency fit with the rest of your stack. Bring your own cloud to keep data inside your own account.

Explore BYOC
Dragonfly 24x7 expert support

Expert support

Backed by the team that built it

Dragonfly Cloud is built, managed, and supported by the Dragonfly team. Get help direct from the engineers behind the engine, backed by contractual SLAs.

See support
Dragonfly Cloud management console: dashboards, logs, and alerts

Dev tools & monitoring

Observability built in

A management console gives you dashboards, graph visualization, logs, alerts, and API integration out of the box, plus automatic failover with multi-region backups.

See monitoring
Dragonfly Cloud security: RBAC, ACL, and admin audit logs

Security

Enterprise controls for your data

RBAC, ACL, and user access management with admin audit logs keep the right people in control. Bring-your-own-cloud keeps your data inside your own account, ready for security reviews.

Visit trust center

Frequently asked

Questions, answered.

Straight answers on cost, migration, scaling, and support.

How much can Dragonfly Cloud actually save us?+
Dragonfly's shared-nothing, multi-threaded architecture runs the same workload on far less hardware. Teams typically reduce infrastructure costs by around 30% and total costs by over 50%, and up to 80% overall, while paying only for the memory they provision.
Do we have to change our code to move?+
No. Dragonfly Cloud runs the same engine as self-hosted Dragonfly, so your commands, data structures, clients, and libraries stay exactly as they are. You repoint your application at the Cloud endpoint.
Who owns and controls the data?+
With bring-your-own-cloud, your data stays inside your own cloud account in the region you choose. Access is governed by RBAC and ACL, and admin audit logs record administrative activity.
How does scaling work?+
You manage memory capacity, not instances or clusters. Autoscaling handles traffic spikes, and Swarm shards automatically across hundreds of TB and hundreds of millions of RPS, so there is no manual sharding or rebalancing to build.
What support do we get?+
You get support directly from the Dragonfly engineering team, backed by contractual SLAs, plus automatic failover, multi-region backups, and managed patching and upgrades.

Try Dragonfly Cloud free.

Spin up a fully managed instance in minutes.

Try it free