Choose a database for your Flutter app

Flutter doesn't include a built-in database, so you can choose one based on the shape of your data and where it needs to live. For most Flutter apps, start with one of these options:

• Cloud Firestore for scalable NoSQL data that syncs across devices in realtime.
• Firebase SQL Connect for relational data that needs a structured SQL schema.
• On-device SQLite or key-value storage for local data that doesn't need a cloud backend by default.

Here's a quick look at how each option stores and syncs data:

Cloud Firestore is the recommended default when your app needs a scalable cloud database that keeps data in sync across users and devices—such as in chat apps, collaborative tools, live multiplayer games, or offline-first field service and inspection apps. It stores data as documents organized into collections, and the cloud_firestore package lets your Flutter app read and write data directly from client code, secured with Cloud Firestore Security Rules.

Cloud Firestore is a good fit when your app needs:

• Realtime sync: listen to a document or query and rebuild your UI whenever the data changes, for example, with a StreamBuilder widget.
• Offline support: read and write cached data while the device is offline, and automatically sync changes when the connection returns, without building a custom offline queue. Offline persistence is enabled by default on Android and Apple platforms, and can be enabled on the web.
• Flexible, evolving data: store nested or semi-structured data without defining a schema up front.
• Automatic scaling: handle growing traffic without provisioning or managing servers.

To add Cloud Firestore to your project, run:

$ flutter pub add cloud_firestore

Then follow the Cloud Firestore quickstart to set up your database.

For structured, relational data—such as in an ecommerce storefront, appointment booking system, personal finance tracker, or inventory system—Firebase SQL Connect (formerly Firebase Data Connect) is the recommended default. SQL Connect is backed by a fully managed PostgreSQL database on Cloud SQL. You define your schema, queries, and mutations in GraphQL, deploy them to the server, and call them from your Flutter app using a generated, type-safe Dart SDK powered by the firebase_data_connect package.

Key capabilities include:

• Relational data: model entities with relationships, such as users, orders, and products, and query across them.
• A strict schema: enforce data types, relational constraints, and foreign keys in the database.
• Type-safe generated SDKs: call generated Dart methods for each predefined server query and mutation instead of writing raw queries or duplicating data models by hand. Because SQL Connect generates client SDKs directly from your backend schema, your Flutter app stays in sync with your database without cross-language schema drift.
• Realtime updates: subscribe to queries for automatic or configured refresh updates when data changes.
• Optional query caching: cache query responses on the client to reduce network requests and read cached results while offline.
• Vector search: build AI-powered features, such as semantic search, on top of your relational data.

To add the SQL Connect plugin to your project, run:

$ flutter pub add firebase_data_connect

Next, follow the guide to use generated Flutter SDKs with SQL Connect.

If your data doesn't need to leave the device, you can store it locally without a cloud backend:

• SQLite: store and query larger amounts of structured relational data on the device. Use sqflite for direct SQL queries, or drift for type-safe Dart queries, reactive data streams, and schema migrations across mobile, desktop, and web. For a step-by-step example, check out Persist data with SQLite, or read Persistent storage architecture: SQL to learn how to structure a local database layer in a larger app.
• Key-value storage: save small amounts of simple data, such as user preferences and settings, with the shared_preferences package. Get started with the Store key-value data on disk recipe, or check out Persistent storage architecture: Key-value data. To store sensitive data like authentication tokens or encrypted records, use flutter_secure_storage instead.

By default, on-device SQLite and key-value packages store data only on the local device, so data isn't synced across devices or preserved if a user uninstalls your app without a backup or sync service. If users expect automatic access to their data across devices or after switching phones, use a cloud database like Cloud Firestore or Firebase SQL Connect instead.

You don't have to pick just one option. For example, your app might use Cloud Firestore or Firebase SQL Connect for shared, synced data, alongside shared_preferences for device-specific settings like a dark mode toggle. To learn how to combine local and remote data sources so your app keeps working without a network connection, check out Offline-first support.

Both Cloud Firestore and Firebase SQL Connect also work alongside other Firebase services, including Firebase Authentication (firebase_auth), Cloud Storage for Firebase (firebase_storage), and Firebase AI Logic (firebase_ai). For example, your app can capture receipt or inspection photos, upload them to Cloud Storage, extract structured fields with Firebase AI Logic, and sync the resulting records in Cloud Firestore or SQL Connect—all within a single project.

If your app needs custom backend logic, you can also write Cloud Functions for Firebase or backend servers in Dart and share a single Dart package for your data models, serialization, and validation logic across both your Flutter client and backend.

To start building with a cloud database, check out the following resources:

• Add Firebase to your Flutter app
• Firebase for Flutter
• Get started with Cloud Firestore
• Firebase SQL Connect for Flutter