Realistic data, without a database.

An API endpoint that returns deterministic JSON sampled from a curated pool of identities, companies, posts, products, and reviews. Cross-references resolve. Same seed returns the same bytes. No schema, no migrations, no setup.

$curl -H "Authorization: Bearer sk_live_…" app.seedkit.dev/api/v1/data/identities
Read the API reference →
Chapter 01The pool

Hand-curated, internally coherent.

Five datasets ship with seedkit and grow release over release: ~540 identities, ~580 companies, ~300 blog posts, ~220 products, ~500 product reviews. Each record is realistic on its own — diverse locales, industries, and voices — and the records reference each other consistently.

Every posts.author_id resolves to a real identities.id. Every reviews.product_id resolves to a real products.id. Every products.company_id resolves to a real companies.id. You can stitch them into your own UI without dangling references.

Records are tagged with industry, locale, and free-form domains tags so you can pull a slice that fits your scenario — a healthcare demo, a fintech landing page, a multilingual content fixture.

datasets snapshot
DATASET COUNT EXAMPLE
identities ~540 maya-okonkwo, dr-james-park
companies ~580 aventar-health, kite-labs
posts ~300 shipping-ml-on-edge
products ~220 aventar-vitals-monitor
reviews ~500 vitals-monitor-mr3a
 
// FK guarantees:
posts.author_id → identities.id // always resolves
reviews.product_id → products.id // always resolves
products.company_id→ companies.id // always resolves
Chapter 02One HTTP call

Curl in. JSON out. Stable across runs.

Authenticate with an sk_live_… PAT (the same token the CLI uses) and hit /api/v1/data/<dataset>. Up to 100 records per call. Filter by industry, locale, domain, or by foreign-key relationships (author_id, product_id, company_id).

Pass ?seed=anything and the response is byte-stable: same dataset + same filters + same seed → same items, same order, every time. Safe to bake into snapshot tests, demo recordings, or shared fixtures across your team.

Need just one record by id? Hit /api/v1/data/<dataset>/<id>. Useful for resolving foreign keys returned by the list endpoint.

$ curl -H 'Authorization: Bearer sk_live_…' \…
// sample 5 healthcare identities, deterministically
$ curl -H "Authorization: Bearer sk_live_…" \
  "app.seedkit.dev/api/v1/data/identities?\
    industry=healthcare&count=5&seed=demo"
 
HTTP/1.1 200 · application/json
{
  "data": [
    { "id": "maya-okonkwo", "job_title": "Pediatric Nurse", … },
    { "id": "dr-james-park", "job_title": "Cardiologist", … },
    … 3 more
  ],
  "count": 5,
  "total_available": 33,
  "seed": "demo"
}
Chapter 03When to reach for this

Skip the schema. Or don't — the choice is the point.

The Synthetic Data API is for when you want realistic data now and don't want to think about a database. Pulling identities for a frontend prototype, dropping reviews into a Figma mock-up, populating a hero section with believable copy, building a Storybook fixture, or backing a 30-second product demo — all one curl away.

When you do have a real schema and want data shaped to it (custom tables, your own foreign keys, columns we don't know about), reach for the schema-driven generator instead. Both are deterministic, both are reproducible — they just start from different ends.

Pro plans include 10,000 calls per month; Team includes 100,000; Enterprise is unmetered. Each request counts as one call no matter how many records you fetch — pulling 100 items costs the same as pulling one.

decision matrix
YOU WANT… → USE
data shaped to YOUR schema → generator
realistic JSON, no DB → data API
seed Postgres for tests → generator
identities for a Figma demo → data API
FK-correct rows in 8 tables → generator
posts + reviews for content → data API
 
// both are deterministic. both are reproducible.
// they just start from different ends.

Try it in a curl.

Mint a PAT in Settings → Tokens, then sample the pool. Pro plans include 10K calls/month.

See pricing →API reference →

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