DocumentDB vs Firestore
DocumentDB
documentdb.io
Firestore
cloud.google.comSide by side
| Figure | DocumentDB | Firestore |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | Yes |
| Open source | — | — |
| Web | — | Yes |
| Windows | Yes | — |
| macOS | Yes | — |
| Linux | Yes | — |
| iOS | — | Yes |
| Android | — | Yes |
| API | Yes | Yes |
| Self-hosted | Yes | — |
| Plans published | 1 | 9 |
| Spec score · Document Databases | 8.6 (#5) | 9.0 (#2) |
| Product | DocumentDB is an open-source, MongoDB-compatible document database built on PostgreSQL. | Firestore is a fully managed, serverless document database for building AI, web, and mobile applications. |
| Security | The project lists SCRAM authentication; its gateway documentation says TLS terminates on the gateway port, whose default is 10260. | Firestore integrates with Cloud IAM, Cloud Identity Platform, and Firebase Authentication for identity-based access controls and data validation. |
| Vector search | Vector search for embeddings and similarity search is powered by pgvector. | Yes |
Plans
Firestore
- Standard edition — document readsPrice not published
- Standard edition — document writesPrice not published
- Standard edition — document deletesPrice not published
- Standard edition — stored dataPrice not published
- Free quotaFree
- Free quotaFree
- Standard edition pay-as-you-go$0.03 a month
- Enterprise editionPrice not published
- Standard editionPrice not published
In their own words
DocumentDB
DocumentDB is an MIT-licensed, open-source document database built on PostgreSQL for MongoDB compatibility. It stores native BSON documents, supports document CRUD and PostgreSQL SQL operations, and offers single-field, multi-key, compound, text and geospatial indexes, including nested fields. Pgvector powers vector search for embeddings and similarity search. MongoDB-compatible clients and tools include mongosh, PyMongo, the official MongoDB Node.js driver and a VS Code extension. The extension can create a local database, generate credentials and save a connection. DocumentDB runs locally in Docker or on a user’s infrastructure or cloud. Its gateway supports SCRAM-SHA-256 authentication, with TLS terminating at default port 10260. Hosted packages currently cover Ubuntu 24.04 and EL9 with PostgreSQL 17 or 18 for amd64 and arm64. A Kubernetes operator for failover, replication, backups and TLS is in preview and requires Kubernetes 1.35+ and cert-manager.
Firestore
Firestore is a fully managed, serverless document database for building AI, web, and mobile applications. It stores flexible JSON data and supports vector search, full-text search, geospatial queries, and ACID transactions on JSON or BSON-compatible documents. Client SDKs for Android, iOS, and web provide real-time updates and automatic syncing for offline access. Firestore can replicate data across multiple regions with strong consistency; multi-region deployments have a stated 99.999% SLA-backed availability. Integrations named for generative AI application work include Google AI Studio, remote MCP, LangChain, and LlamaIndex. Identity controls connect with Cloud IAM, Cloud Identity Platform, and Firebase Authentication. Data is encrypted before it is written to disk, with customer-managed keys available through Cloud KMS. The free quota applies to one database per project and includes 1 GiB of storage, daily read, write, and delete allowances, and 10 GiB monthly outbound transfer. New customers receive $300 in credits, separate from that quota. Server client libraries bypass Firestore Security Rules, so server access needs IAM controls.