Diagram Any Neon Branch, Including pgvector Columns

Neon is serverless PostgreSQL with database branching: fork production into an isolated branch in seconds, each branch with its own connection string. DbSchema connects with that string and turns the branch's schema — pgvector columns included — into an ER diagram.

Paste a branch's connection string, get that branch's schema as a diagram
Compare schemas between branches before a migration merges
Full PostgreSQL SQL in the editor — CTEs, window functions, extensions
Inspect pgvector embedding columns while debugging AI pipelines

available for

ISO-27001 certified

Free Community Edition No signup required 15-day Architect trial included

Free Download

DbSchema diagramming a Neon branch-specific PostgreSQL schema

Trusted by teams at

Branches Fork in Seconds
Schema Drift Follows Right Behind

Neon separates compute from storage, so a branch costs seconds and every developer or CI run can have one — auto-suspending when idle. Apply migrations independently for a week, and "what exactly is on this branch?" starts needing a real answer.

DbSchema answers it visually: connect to the branch endpoint, see its schema as a diagram, and set it against main before the merge.

Connect, Compare & Verify
Every Branch You Fork

PostgreSQL tooling that understands a per-branch world.

Reverse engineer a Neon branch schema into an ER diagram

One Branch, One Diagram

Every Neon branch exposes its own endpoint; point DbSchema at it and that branch's tables, keys, and relationships appear as an ER diagram.

Compare Neon branch schemas before merging a migration

Diff Branches Before Merging

Compare a feature branch's schema against main, walk through every structural difference, and know what a migration really changed before it travels on.

Inspect pgvector embedding columns in the DbSchema data explorer

See the Embeddings Land

The data explorer renders pgvector columns as array-like values, so RAG pipelines can be checked visually — with filters and foreign-key navigation for everything else.

Neon Branches the Data
Git Versions the Design

DbSchema keeps the schema in a model file, so the design gets a history of its own: commit it beside the application code, review schema changes in pull requests, and regenerate the diagram for whichever Neon branch you connect to next.

Schema synchronization between a Git-versioned model and Neon branches

Read Git Workflow Docs

Neon Connection Facts

Neon speaks the PostgreSQL wire protocol, so DbSchema's bundled PostgreSQL driver (org.postgresql.Driver) connects out of the box — the workflow on our PostgreSQL page applies to Neon unchanged. Copy the branch's connection string from the Neon dashboard; as a JDBC URL it reads jdbc:postgresql://ep-xxx.us-east-2.aws.neon.tech/mydb?sslmode=require, where the hostname encodes the compute endpoint ID and AWS region. SSL (sslmode=require) is mandatory on every Neon connection.

For applications making many short-lived connections, Neon offers a pooler endpoint whose hostname ends in -pooler.us-east-2.aws.neon.tech; it avoids exhausting the connection limit on small compute tiers. Idle branches auto-suspend and wake when DbSchema connects. Since each branch has its own endpoint, save one connection profile per branch. Driver details: Neon JDBC driver page.

Diagram the Branch Before the Pull Request

No signup — the installer carries the free Community Edition and a 15-day Architect trial.

Download Free

Frequently asked questions

Yes. DbSchema connects to any Neon branch independently using that branch's own connection string, letting you reverse-engineer and visualize the schema specific to that branch.

Neon uses the standard PostgreSQL JDBC driver with a URL such as jdbc:postgresql://ep-xxx.us-east-2.aws.neon.tech/mydb?sslmode=require, and SSL is mandatory for every Neon connection.

Yes. The data explorer renders Neon's vector columns, added by the pgvector extension, as array-like values so you can confirm embedding dimensions were populated correctly.

Teams working with Neon often use these engines too. Explore dedicated guides and JDBC setup for each.