AWS Amplify DataStore

AWS Amplify DataStore is an on-device repository for applications that read, write, and observe data online or offline. Data can stay local or synchronize bidirectionally with the cloud and across devices. GraphQL data models can be converted for JavaScript, iOS, and Android applications; the DataStore API supports saving, querying, updating, deleting, and observing models. With cloud sync enabled, the Amplify CLI can deploy an AWS AppSync backend with DynamoDB tables for models and a Delta Sync table. DataStore includes SQLite and IndexedDB storage adapters and supports Auto Merge, Optimistic Concurrency, and custom Lambda conflict resolution. Mobile support includes iOS, Android, React Native, and Flutter, with SDKs for JavaScript, Swift, Java, and Dart. Listed service capabilities also include edge delivery, serverless and Node.js runtimes, preview deployments, 1000 min/mo of build minutes, and 15 GB/mo of bandwidth. The former DataStore page redirects to Amplify Data, and AWS documentation says DataStore will receive no new features and provides migration guidance. Pricing starts at $0.01; there is no free plan. AWS says backend resources have no monthly minimum and are charged based on use.

DataStore provides offline storage, synchronization, and conflict-handling options for application data. AWS's migration guidance and no-new-features status are important considerations before adopting it.

DataStore may suit developers building applications that need local offline data and optional cloud synchronization. It is important to consider AWS's migration guidance and the statement that the product will receive no new features.

AWS Amplify DataStore is an on-device data layer for applications that must read, write, observe, and synchronize records across offline and connected states. It suits teams building mobile or web clients around GraphQL models who need local persistence and managed AWS synchronization. The strongest case is offline-first application data with controlled conflict behavior; the main caution is that AWS now directs developers toward a replacement path rather than adding new DataStore capabilities.

Overview

DataStore turns GraphQL data models into application models for JavaScript, Swift, Java, or Dart clients. Generated models can be saved, queried, changed, removed, or observed through the DataStore API, giving developers one local repository for normal CRUD work and live data updates.

It can run entirely offline and locally without an AWS account. With cloud synchronization enabled, local records synchronize bidirectionally with the backend and across devices. The Amplify CLI can deploy an AWS AppSync backend, DynamoDB tables for each model, and a Delta Sync table to support synchronization.

Key features

Offline storage and synchronization

Persistent on-device storage keeps an app usable when connectivity is intermittent, then synchronizes changes when configured for the cloud. SQLite and IndexedDB adapters cover the shipped storage implementations, making the approach practical for mobile and browser clients rather than requiring a separate local database design.

Observation and conflict handling

The observe and observeQuery APIs provide live updates to application data. For synchronized records, Delta Sync and three conflict strategies—Auto Merge, Optimistic Concurrency, and custom Lambda resolution—let a team choose between convenience, stricter collision detection, or application-specific rules. That flexibility matters for collaborative or multi-device workflows, but custom resolution adds backend design and maintenance.

Authorization and current Amplify Data integrations

AWS Amplify Data describes fine-grained authorization for authenticated users, owners, and groups, with permissions assignable at model or field level. It also enables real-time use cases by default, allowing frontend clients to subscribe to updates. The current service uses AWS AppSync and Amazon DynamoDB, and can integrate with third-party PostgreSQL providers and Amazon Bedrock.

Pricing

DataStore is a paid offering from $0.01. There is no free plan. AWS Amplify pricing uses consumption-based billing for backend resources, with no monthly minimum; AppSync and DynamoDB are examples of resources charged by use. That structure suits prototypes and workloads with variable traffic, but production cost depends on synchronized data and the backend resources your application consumes rather than a single all-inclusive DataStore subscription.

Local-only operation without an AWS account avoids cloud synchronization costs and is the sensible fit for apps that only need device-local persistence. Teams needing shared, cloud-backed records should budget for AppSync and DynamoDB usage and monitor those services separately.

Platforms

DataStore supports Android, iOS, and web, with mobile coverage including React Native and Flutter. SDK languages are JavaScript, Swift, Java, and Dart. The broader Amplify environment lists edge delivery, a serverless runtime, a Node.js runtime, 1,000 build minutes per month, 15 GB monthly bandwidth, and preview deployments, useful context for teams considering Amplify as a wider application platform rather than only a data layer.

Who it's for

Choose DataStore for an offline-first mobile or web product whose records need automatic two-way synchronization, generated models, subscriptions, and explicit conflict policies. It is especially suitable when your team already uses AppSync, DynamoDB, or the Amplify toolchain and wants local and cloud data behavior connected.

It is a poor fit for a new project that wants an actively expanding data product with a long forward roadmap. AWS documentation says DataStore will receive no new features and provides migration guidance to replace it with Apollo Client and Amplify subscriptions. That lifecycle status should outweigh convenience for teams starting a long-lived platform.

Pros and cons

• Pro: Offline persistence works locally without an AWS account, so disconnected use cases do not require cloud infrastructure.
• Pro: Bidirectional synchronization, Delta Sync, and selectable conflict strategies address multi-device data without forcing one collision policy.
• Pro: Generated GraphQL models and observe APIs reduce the amount of client-side data plumbing for supported SDKs.
• Con: There is no free plan, and backend usage is billed separately according to consumption.
• Con: AWS has frozen DataStore feature development and documents migration away from it, creating roadmap risk for new adopters.
• Con: Custom Lambda conflict resolution can deliver precise business behavior but requires additional implementation and operations work.

Alternatives

Firebase Cloud Messaging is the better choice when the requirement is no-cost messaging delivery across Android, iOS, web, API, and extension platforms rather than an offline synchronized data repository.

Railway fits teams wanting a freemium deployment platform with a trial that includes a one-time $5 credit, two replicas, 1 GB RAM, and 2 vCPU, instead of DataStore's application-data synchronization model.

sh0 is worth choosing when a free self-hosted stack with unlimited apps and deployments is more important than managed AppSync synchronization; its Pro plan is $19.00 per month.

Netlify suits web teams seeking a free tier with 300 credits per month and one concurrent build, or a $9.00-per-month Personal plan, rather than a client data layer.

Temps is an alternative for teams preferring self-hosted infrastructure and local, unmetered telemetry on their own hardware.

Cloudflare Pages is a fit for web delivery with a free plan that includes 500 builds per month and one concurrent build.

Easel targets edge and serverless workloads, with a free Hobby tier including 1 million edge requests per month.

Encore is an alternative for teams wanting a free Starter tier with two development environments and 1 million trace events per month.

Preview Environment Software is the relevant category when temporary deployment environments, not synchronized client data, are the priority. For broader mobile data tooling, compare the Mobile Databases directory.

Verdict

DataStore remains a capable choice for an existing Amplify team that needs offline storage, two-way synchronization, generated models, and configurable conflict handling now. The decisive reason to look elsewhere is its frozen feature roadmap and AWS migration guidance: new, long-lived projects should evaluate Apollo Client with Amplify subscriptions or another actively developed data architecture before committing.