What Is Upstash?
Upstash is a serverless data platform intended for developers building real-time and event-driven applications. It provides managed data stores (i.e., Redis) and a messaging platform (QStash) with flexible pricing options. This makes it appropriate for cloud-native environments, including serverless platforms like AWS Lambda, Vercel, and Cloudflare Workers.
Upstash is designed for ease of use without the need to manage infrastructure. It aims to eliminate provisioning and scaling complexities by automatically adjusting resources based on workload. Developers interact with Upstash services using APIs, and its HTTP-based interfaces enable integration across various environments.
Key Features of Upstash
Here are the key features that make Upstash a possible choice for developers building serverless and real-time applications:
- Serverless architecture: Upstash services automatically scale with traffic, helping avoid manual capacity planning or server management.
- Pay-per-request pricing: Instead of traditional instance-based billing, Upstash charges based on usage, making it suitable for applications with variable workloads.
- Multi-region deployment: Data can be replicated across multiple regions to enable availability for global users.
- HTTP-based APIs: Unlike typical Redis services that use the TCP protocol, Upstash offers HTTP APIs that support integration with serverless and edge functions.
- Persistent storage: Upstash provides durable storage for Redis data, aiming to ensure resilience across restarts and deployments.
- Role-based access control: Permissions can be assigned to users or applications to secure access to datasets.
- Built-in caching and TTL: Support for time-to-live (TTL) and caching mechanisms helps optimize data lifecycle management and performance.
Core Services of Upstash
Upstash Redis
Upstash Redis is a serverless, Redis-compatible database designed for low-latency, scalable applications. Key features include:
- High availability: 99.99% uptime with automatic scaling.
- Global replication: Data is replicated across multiple regions to help reduce latency and improve availability.
- Persistent storage: Provides data durability across restarts and deployments.
- Security: Supports TLS encryption and role-based access control.
- HTTP-based API: Provides an HTTP interface, suitable for serverless and edge environments.
- Redis API compatibility: Supports most Redis commands up to version 6.2, with ongoing additions from newer versions.
Upstash Vector
Upstash Vector is a serverless vector database for storing and querying high-dimensional vector embeddings. Key features include:
- Approximate nearest neighbor (ANN) search: Uses the DiskANN and FreshDiskANN algorithms with improvements for similarity searches.
- Serverless architecture: Does not require infrastructure management, scaling automatically based on usage.
- Indexing options: Supports dense, sparse, or hybrid indexes with configurable similarity functions like cosine similarity.
- HTTP API and SDKs: Provides HTTP-based APIs and SDKs for Python, TypeScript, and other languages.
- Namespace support: Allows partitioning of an index into multiple isolated namespaces.
QStash
QStash is a serverless messaging and scheduling solution intended for message delivery in serverless architectures. Key features include:
- Push-based messaging: Delivers messages to API endpoints without requiring long-running consumers, which is also serverless friendly.
- Automatic retries: Failed messages are retried automatically, with support for Dead Letter Queues (DLQs) to handle persistent failures.
- Scheduling: Supports cron-based scheduling for recurring tasks.
- Per-request pricing: Charges based on usage, intended for applications with variable workloads.
- Platform agnostic: Works with various publicly accessible HTTP APIs.
Upstash Workflow
Upstash Workflow is a serverless orchestration engine built on top of QStash, enabling the creation of durable workflows. Key features include:
- Step functions: Define workflows as a series of steps that can run independently with automatic retries in case of failure.
- Long-running executions: Supports long-running tasks by breaking them into resumable steps.
- Event-driven: Workflows can wait for external events before proceeding, suitable for user confirmations and asynchronous processes.
- Scheduling and delays: Supports scheduled jobs and long delays, enabling workflows to pause for extended periods.
- Flow control: Supports configuring rate limits and parallelism to prevent overwhelming downstream services.
- Agents API: Can integrate AI agents into workflows for autonomous task execution.
Upstash Pricing Details
Here is an example of Upstash pricing for Redis as of July 2025. Consult the official pricing page to see pricing details for Vector, QStash, Workflow, and Search options.
Upstash Redis Pricing
Free Tier
The free plan is suitable for prototypes and small-scale applications. It includes:
- Cost: $0/month
- Data size: 256 MB
- Monthly commands: 500,000
- Use case: Suitable for personal or experimental projects
Pay As You Go
This plan provides flexible, usage-based pricing:
- Cost: $0.20 per 100,000 commands
- Data size: Up to 100 GB
- Monthly bandwidth: Unlimited
- Use case: Suitable for applications with variable traffic patterns
Fixed Plans
For workloads with predictable usage, Upstash offers fixed plans with between 250MB and 500GB of storage. Here is the pricing for the smallest plan with 250MB:
- Cost: $10/month (plus $5 per additional read region)
- Data size: 250 MB
- Monthly bandwidth: 50 GB
- Use case: Consistent loads with predictable pricing
Prod Pack
The Prod Pack is designed for production use and includes advanced features:
- Cost: +$200 per database
- Features:
- Uptime SLA
- Role-Based Access Control (RBAC)
- Encryption at rest
- SOC-2 compliance
- Integration with monitoring tools like Prometheus and Datadog
Enterprise
The Enterprise tier is for large-scale and mission-critical systems:
- Custom pricing (contact sales)
- Includes everything in the Prod Pack, plus:
- Support for over 100,000 commands per second
- Unlimited bandwidth and database instances
- SLA-backed professional support
- Dedicated resources for isolation
- HIPAA compliance
Upstash vs. Redis Cloud vs. Dragonfly Cloud
Redis Cloud, Dragonfly Cloud, and Upstash are all cloud-based services that provide managed Redis instances, but they cater to different use cases and audiences. Below is a comprehensive comparison of the three.
Performance, Scalability, and Target Audience
Redis Cloud
Redis is the de facto choice for many teams. It has tenure and is generally considered stable. It’s a “safe choice” for many teams. Redis Cloud, the managed version of Redis, can automatically adjust resources based on load, which ensures that performance remains consistent even under varying loads. It employs advanced clustering technology for seamless scaling without downtime.
Redis Cloud’s performance, scalability, and cost efficiency are typically limited by Redis itself. A standalone Redis instance cannot utilize multiple CPU cores, and premature horizontal scaling is often required. For example, if you try to provision a 310GB cache in Redis Cloud, it will automatically deploy 13 Large instances, which are actually very small, with only 25GB of memory each. This shows the limitations of Redis, which is single-threaded and unable to utilize modern cloud servers.
Dragonfly Cloud
The Dragonfly project is designed for the most demanding workloads, exceeding the performance of Redis. It offers high throughput and low latency by leveraging a multi-threaded shared-nothing architecture optimized for modern hardware. Dragonfly supports large numbers of concurrent requests and scales efficiently, making it suitable for both high-traffic applications and compute-intensive operations.
Dragonfly Cloud aims to be the most performant in-memory key-value data store service. It achieves this by scaling vertically first (standalone with replicas on powerful multi-core servers) and then horizontally if necessary (with Dragonfly Swarm multi-node clusters).
You can start small on Dragonfly Cloud, but expect it to handle millions of ops/sec or even practically unlimited performance (100+ million ops/sec, 100TB+ memory) when your application goes viral.
Upstash
Upstash is a serverless platform designed to scale seamlessly in theory. It provides good performance at a small scale but is designed with a serverless mindset, making it particularly suitable for applications with variable workloads. It automatically scales down to zero when not in use, which can be cost-effective for intermittent workloads.
However, Upstash is probably not ideal for large-scale or demanding applications. At the time of writing, their homepage reports handling 40 billion Redis commands per week, which averages around 66,138 commands per second across the entire platform. This suggests that the platform is mainly used by small and medium workloads.
Upstash favors HTTP-based client SDKs (instead of TCP for many databases), which is dictated by its serverless nature. This requires authentication for every request, which reduces performance.
While the serverless model excels in scalability and ease of use for moderate workloads, its design is not optimal for highly demanding, low-latency use cases. The serverless model also doesn’t cater well to scenarios where persistent connections and high throughput are critical.
Pricing Model
Redis Cloud
Redis uses a tiered pricing model based on the size of the data set and the level of performance needed. This can include dedicated instances for higher performance requirements.
Due to the limitations of Redis, the service provisions a large number of instances when scaling up. For example, their pricing tool shows that if you scale up to 310 GB, it will deploy 13 Large instances, each handling 25GB. This shows that Redis Cloud can get really expensive when scaling up dataset size, because they need to manage each Redis instance within minimal hardware specs.
Dragonfly Cloud
Dragonfly Cloud has a simplified model where price is mainly determined by memory size and compute tier. It offers simplified and transparent pricing based on instance size and usage. While not as granular as Upstash’s per-request billing, it remains cost-competitive, especially for users seeking high performance without the overhead of Redis enterprise features.
Dragonfly Cloud Pricing
Dragonfly Cloud is very transparent about deployment costs: if you’re under 600GB, you can choose to stay with the standalone mode. If you need bigger data stores, you can choose the Dragonfly Swarm cluster mode. This makes it a viable option for mid-sized projects that need performance predictability without complex pricing tiers. The price is primarily determined per GB of memory. Note that a 100GB Dragonfly Cloud data store with the Standard compute tier can typically handle ~1 million operations per second.
Upstash
Upstash has flexible pricing for pay-as-you-go and dedicated/fixed pricing as well. It charges based on the number of requests and the amount of data stored and transferred. This can be more cost-effective for applications with fluctuating usage patterns or lower throughput needs.
However, the solution can get very expensive for large-scale use cases. For example, the fixed price for 100GB is $800 per month with a rather small requests/second limit of 16,000.
Features and Capabilities
Redis Cloud
Redis Cloud focuses on innovating on enterprise and peripheral features (i.e., active-active, data tiering). It provides a range of features, including support for all Redis data types, advanced security features, integrated billing, and 24/7 support. It also offers the Redis on Flash technology, which allows storing large data sets in SSDs.
Dragonfly Cloud
Dragonfly functions much like traditional Redis-based solutions but delivers significantly higher performance. It supports standard key-value caching, JSON operations, and search capabilities.
Emphasizing performance and operational simplicity, Dragonfly Cloud supports most Redis commands and data types and includes built-in observability and snapshot-based persistence for fast restarts. Dragonfly is developer-friendly, offering an intuitive interface and minimal configuration.
Users can power messaging or background job processing systems like BullMQ and Sidekiq with Dragonfly Cloud as the back-end storage. It functions as a high-performance, drop-in replacement for Redis and Memcached while maintaining high API compatibility.
Upstash
Upstash focuses on simplicity and ease of use, with features like instant setup, automatic scaling, and a straightforward dashboard for managing instances. It supports HTTP-based API, which can be beneficial for integrating with serverless architectures and event-driven applications.
The solution offers more than just Redis—it provides additional services like Vector (for embeddings), QStash (a message queue), Workflow (for background jobs and workflows), and Search. Upstash structures these as standalone products, allowing developers to pick specific solutions without tightly coupling workloads.
Separating data across services introduces some trade-offs, but this approach can make Upstash a more versatile, modular platform.
Use Cases
Redis Cloud
Suitable for businesses and projects that require robust, scalable, and secure Redis instances. It’s well-suited for real-time analytics, high-speed transactions, and session caching for applications.
Dragonfly Cloud
Suited for teams that handle the most demanding workloads with less operational overhead and more predictable performance. It’s ideal for caching layers, real-time analytics, and pub/sub systems where latency and throughput are critical. If teams don’t use features like active-active replication (which at the time of writing Dragonfly doesn’t offer yet), migrating from other services to Dragonfly Cloud generally guarantees cost reduction. When the code is more efficient, less hardware is required—it is that simple and true for all kinds of software.
Upstash
Best suited for developers and startups looking for an easy-to-use, cost-effective solution for smaller projects or applications with variable traffic. It’s particularly useful for serverless applications, microservices architectures, and event-driven systems.
Dragonfly: The Next-Generation In-Memory Data Store
Dragonfly is a modern, source-available, multi-threaded, Redis-compatible in-memory data store that stands out by delivering unmatched performance and efficiency. Designed from the ground up to disrupt existing legacy technologies, Dragonfly redefines what an in-memory data store can achieve. With Dragonfly, you get the familiar API of Redis without the performance bottlenecks, making it an essential tool for modern cloud architectures aiming for peak performance and cost savings. Migrating from Redis to Dragonfly requires zero or minimal code changes.
Key Advancements of Dragonfly
- Multi-Threaded Architecture: Efficiently leverages modern multi-core processors to maximize throughput and minimize latency.
- Unmatched Performance: Achieves 25x better performance than Redis, ensuring your applications run with extremely high throughput and consistent latency.
- Cost Efficiency: Reduces hardware and operational costs without sacrificing performance, making it an ideal choice for budget-conscious enterprises.
- Redis API Compatibility: Offers seamless integration with existing Redis applications and frameworks while overcoming its limitations.
- Innovative Design: Built to scale vertically and horizontally, providing a robust solution for rapidly growing data needs.
Dragonfly Cloud is a fully managed service from the creators of Dragonfly, handling all operations and delivering effortless scaling so you can focus on what matters without worrying about in-memory data infrastructure anymore.
