DBSprout

Why DBSprout?

Schema-First

Point at your database, DDL file, DBML, Prisma schema, or Mermaid diagram. DBSprout reads your schema and generates matching data automatically.

Offline by Default

Ships with an embedded AI model. No internet, no API key, no account needed. Works anywhere — CI pipelines, air-gapped environments, your laptop on a plane.

Deterministic Output

Same seed produces identical output every run. Perfect for CI/CD pipelines, reproducible test fixtures, and team collaboration.

Blazing Fast

100K+ rows/sec with heuristic generation. PostgreSQL COPY and MySQL LOAD DATA for bulk insertion at database-native speeds.

Foreign Key Aware

Automatically resolves FK dependencies via topological sort. Handles cycles, self-references, and junction tables out of the box.

Privacy Gradient

Four privacy tiers — Local, Redacted, Cloud, Training — each with full audit logging. You control what data leaves your machine.

Quick Example

# Install
pip install dbsprout

# Point at your database
dbsprout init --db postgresql://localhost/myapp

# Generate 1000 rows per table
dbsprout generate --rows 1000

# Or use a schema file directly
dbsprout init --file schema.sql
dbsprout generate --output-format csv

Supported Databases & Formats

Input	Output
PostgreSQL, MySQL, SQLite, MSSQL	SQL INSERT
SQL DDL files	CSV
DBML, Prisma, Django models	JSON
Mermaid ER, PlantUML	Parquet
MongoDB (planned)	PostgreSQL COPY
	MySQL LOAD DATA