DocumentDB claims 85% faster reads than MongoDB by using a Postgres storage layer

A new open-source database offers MongoDB API compatibility on a Rust and PostgreSQL foundation. We analyze its launch claims of superior performance and storage efficiency against the incumbent.

THE ANSWER UP FRONT

DocumentDB is for engineering teams committed to the PostgreSQL ecosystem who need to support a service that requires a MongoDB-compatible API. It offers a path to database consolidation. Teams deeply invested in MongoDB-native features, sharding, or those requiring a decade of production hardening should skip it for now. The bottom line: DocumentDB presents a compelling architectural alternative to MongoDB, but its significant performance and storage claims are founder-reported and require independent verification before any production use.

METHODOLOGY

This is a v0 review based on the project's public launch materials. Our analysis draws exclusively from the founder's published claims and technical descriptions on the DocumentDB website. Independent benchmarks are pending.

• Tool: DocumentDB
• Version: Not specified in launch materials
• Date Observed: June 29, 2026
• Source Signal: The official launch website, https://documentdb.io/, posted by user 'amai'.
• What's Covered: This review covers the tool's stated architecture (Rust query engine, PostgreSQL storage), its core value proposition (MongoDB compatibility), and its specific, unverified claims regarding performance and storage reduction.
• What's Not Covered: We have not conducted independent performance benchmarks, tested the limits of its MongoDB API compatibility with complex queries or aggregation pipelines, or assessed its long-term stability under production workloads. This analysis does not compare DocumentDB to other Postgres-based MongoDB alternatives like FerretDB.

WHAT IT DOES

DocumentDB positions itself as a drop-in replacement for MongoDB that uses PostgreSQL for its storage backend. The architecture is its main feature.

A MongoDB API on a Postgres foundation

Instead of building a new storage engine from scratch, DocumentDB translates MongoDB wire protocol and query language (MQL) requests into SQL. These SQL queries then run against a standard PostgreSQL database. The goal is to combine the developer experience of a document API with the reliability, maturity, and rich feature set of Postgres. This allows teams to leverage existing Postgres tooling for backups, monitoring, and administration.

Rust-based query engine

The translation layer that sits between the application and PostgreSQL is written in Rust. The choice of Rust suggests a focus on performance and memory safety, aiming to minimize overhead in the query translation and execution path. The website doesn't provide deep technical details on the engine's internals but presents this as a key reason for its claimed performance advantages.

Claims of major efficiency gains

The two headline claims are dramatic. The project's website asserts it is "up to 85% faster" than MongoDB and uses "up to 92% less storage." The storage savings are attributed to PostgreSQL's more compact data representation (likely JSONB) compared to MongoDB's BSON format, which can include significant padding. The performance claim is less specific about the workloads tested but implies that Postgres's mature query planner can outperform MongoDB's on certain read patterns.

WHAT'S INTERESTING / WHAT'S NOT

The most interesting aspect of DocumentDB is its bet on architectural consolidation. Many organizations are trying to reduce the number of database technologies they manage. By providing a popular NoSQL API on top of the industry-standard relational database, DocumentDB offers a compelling path for teams to standardize on Postgres without rewriting applications.

The performance and storage claims, while impressive, are also the biggest question mark. The phrases "up to 85%" and "up to 92%" are classic marketing language for best-case scenarios. The website provides no public, reproducible benchmark suite to validate these numbers. We don't know the hardware used, the dataset shape, the query patterns (read-heavy vs. write-heavy), or the specific MongoDB version it was benchmarked against. Without this data, the claims are just assertions.

What's missing is a clear discussion of the trade-offs. What is the performance cost for write-heavy workloads? How does the translation layer handle complex MongoDB aggregation pipelines or operators that don't have a direct, efficient analog in SQL? True MongoDB compatibility is notoriously difficult to achieve, and the edge cases are where projects like this succeed or fail. The launch materials are silent on the current level of API completeness and known limitations.

PRICING

As of June 29, 2026, DocumentDB is an open-source project available under the Apache 2.0 license. It is free to download and self-host. The website does not list any paid tiers, enterprise licenses, or managed cloud offerings.

VERDICT

DocumentDB is a promising new entrant for a specific user: a developer who wants the MongoDB API but would prefer the operational maturity and ecosystem of PostgreSQL. It's a strong architectural premise. However, the project is clearly in its early stages. Adopting it today would be a bet on the roadmap and the team's ability to deliver on its ambitious performance claims and achieve deep API compatibility. For most teams, DocumentDB is a project to watch and benchmark. For new projects already committed to Postgres, it's worth evaluating against using Postgres's native JSONB capabilities directly.

WHAT WE'D TEST NEXT

Our v2 review would require hands-on benchmarking. First, we would attempt to reproduce the headline performance and storage claims using a standard tool like the Yahoo! Cloud Serving Benchmark (YCSB). We would design tests to measure performance across a mix of workloads, from read-heavy analytics queries to write-intensive transactional patterns. Second, we would assess the depth of MongoDB compatibility by running the test suites of several popular open-source applications built on the MERN stack against a DocumentDB instance to identify unsupported operators or behavioral differences. Finally, we would directly compare its performance not just to MongoDB, but also to native PostgreSQL JSONB queries for an equivalent task.

The investor read

DocumentDB is a play on the database consolidation trend, specifically the movement to build more services on top of PostgreSQL. Its direct competitor is FerretDB. The market signal is that the demand for Postgres is strong enough to support an ecosystem of API-compatibility layers. For DocumentDB to be investable, it needs to move beyond being a purely open-source project. A viable path would be a managed cloud offering (a DBaaS) that can prove a significant total cost of ownership advantage over MongoDB Atlas. This requires verifiable, public benchmarks demonstrating the claimed performance and storage efficiencies. Without a clear commercialization strategy and reproducible proof points, it remains a high-risk, high-potential open-source project rather than a venture-scale business.

Pull quote: “The bottom line: DocumentDB presents a compelling architectural alternative to MongoDB, but its significant performance and storage claims are founder-reported and require independent verification before any production use.”