Fake data. Real databases. Zero setup.

Generate realistic, synthetic data from a schema or plain english description. Instantly pipe the data into existing databases, or provision an ephemeral postgres database and JSON API in seconds.

Get started [for free]Read the docs
Try it now
$npx seedkit-cli new
Setup instructions

Seed smarter, ship faster.

Synthetic data, an ephemeral Postgres, and a JSON API for it — all reproducible from the same seed. Provisioned and accessible in seconds.

$ npx seedkit-cli new --prompt "saas crm"
Designing schema... ✓ 8 tables, 14 relations
Generating data... ✓ 14,820 rows
Provisioning Postgres... ✓ ready
Your database is live — expires in 60 min:
postgres://user:••••@ep-cool-leaf-7281.eu.neon.tech/seedkit_a3fd
// Try: psql $(seedkit url)
// Or open in browser: seedkit.dev/db/a3fd

And everything else you'd expect — in the box.

The boring essentials, required for calm data seeding, done right.

  • Real Postgres URL
  • Foreign-key correct
  • Auto-expiring
  • Built for CI/CD
  • Deterministic seeds
  • EU / GDPR-native
  • Free tier
  • Schema-aware

Pick your entry point.

The web app is for sketching, exploring, and sharing. The CLI is for shipping it into your stack. Same engine, same generations, swappable any time.

Web app

From idea to working DB.

Paste a schema or describe an app in plain English. Watch rows appear, query with SQL, share a link with a teammate. Start fast in the Playground.

seedkit dashboard showing active databases and usage
Fig. — DashboardSynthetic data
CLI

Into your existing stack.

Point it at $DATABASE_URL — your local Postgres, your staging, your CI database. One command, no factory code, no broken seeds.

$ npx seedkit-cli new --prompt "saas crm"
[1/5] Authenticating... ✓ ben@seedkit.dev
[2/5] Designing schema... ✓ 8 tables · 14 relations
[3/5] Generating data...
  ├ users        ✓ 240 rows · 0.4s
  ├ accounts     ✓ 480 rows · 0.6s
  ├ contacts     ✓ 1,920 rows · 1.1s
  ├ opportunities ✓ 2,100 rows · 0.8s
  ├ activities   ✓ 8,400 rows · 1.2s
  └ notes, tags  ✓ 1,680 rows · 0.5s
[4/5] Provisioning Neon... ✓ eu-central-1
[5/5] Applying SQL... ✓ 14,820 rows · 2.8s
✓ ready
postgres://****@ep-cool-leaf-7281.eu-central-1.host.tld/seedkit_a3fd
Fig. — One command, your DBInstall the CLI

“I love maintaining our seed files.” (struck through)Nobody like maintaining seed files. We get it. That's why we built seedkit.

The modern alternative is branching from prod — Neon, PlanetScale, Supabase all sell it. Until legal sees real customers in your dev database. PII leaks, GDPR exposure, the inevitable Slack screenshot of a real user's email. Synthetic data is the only path that's realistic and safe.

“It broke again.”

Every migration invalidates your seed. Every schema change means rewriting factories. seed.ts is a graveyard of stale code that nobody wants to own.

“Faker doesn’t know about FKs.”

You end up writing 200 lines of glue code to make orders reference real users and products reference real categories. That's not seeding — that's scaffolding.

“Production isn’t an option.”

Legal said no. Security said no. GDPR said no. Masking is harder than it looks once you have JSON columns and free-text fields. So you're back to faker.name().

Free to try. Cheap to keep.

No sales call, no card to start. Annual billing and feature comparison live on the pricing page.

Free

For tinkering and side projects.

$0forever

1 ephemeral DB · 10k rows / mo

Start free
ProMost popular

For working developers.

$29/ month

5 DBs · 1M rows · 30-day mock TTL

Go Pro
Team

For teams who ship together.

$99/ month

25 DBs · 10M rows · 90-day mock TTL

Start team trial

All plans · cancel any time · prorated · EU-hosted by default

See full pricing & feature comparison

Fake data, real questions.

Where do these databases actually run?

On Neon, in EU regions by default. Real Postgres 16, real connection strings, real psql-compatible. Neon's serverless architecture means your preview DBs scale to zero between queries — that's why we can offer them on the free tier.

Can I use this with my existing database?

Yes — that's the CLI's main job. Point seedkit seed at any Postgres URL and it inserts there. The Neon preview is a separate feature for when you want a throwaway. Use both, neither, or either.

Does it work with Prisma / Drizzle / Kysely / raw SQL?

Yes to all. Seedkit reads your database schema directly, not your ORM — if your ORM can connect to it, we can seed it. We also parse Prisma schema.prisma and Drizzle TS files directly for the “no DB yet” flow.

What about my data — do you store it?

The generated data lives in your preview DB until it expires, then it's deleted. We never see the data your existing database contains; the CLI runs locally against your DB and only sends schema metadata to our generation service. Self-hosted is on the roadmap.

Still wondering? See the full FAQ on the pricing page or read the docs.

Your next database is a sentence away.

$npx seedkit-cli new
// or paste your schema at seedkit.dev →
Try in browser →Read the docs

Free to start · No card required · MIT-licensed CLI