DEV Community

DSP
DSP

Posted on

SynthForge: fake test data with deterministic seeds, custom schemas, and SQL output

#testing #api #python #webdev

Seeding a database with user1, user2, test@test.com works right up until your UI has to render a real-looking name, your demo needs 500 believable products, or your test suite needs the same dataset on every run, not a fresh snowflake each time.

SynthForge is a synthetic data API we built at DSP to fix that. It's live on RapidAPI with a free tier.

Predefined generators

Seven generators, up to 1,000 records per call: person, company, address, product, user, payment-card, and internet, covering names, safe emails, phones, job titles, SKUs, prices, usernames, IPs, UUIDs, and the rest of the usual test-data furniture.

Payment cards deserve a specific mention: numbers come only from the standard public test ranges (Visa 4242…, Mastercard 5555…, Amex, Discover), are Luhn-valid so your validators accept them, and every record is flagged "test_only": true. They will never work for a real purchase, by design.

The headline: bring your own schema

Describe the shape you want using any of 38 field generators, and get back exactly that shape:

curl -X POST https://synthforge.p.rapidapi.com/v1/generate/schema \
  -H "X-RapidAPI-Key: YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"schema": {"id": "uuid", "name": "product_name",
                  "price": "price", "city": "city", "zip": "zip_code"},
       "count": 2, "seed": 7}'
Enter fullscreen mode Exit fullscreen mode 
[{"id": "6513270e-269e-4d37-b2a7-4de452e6b438", "name": "Premium Lamp",
  "price": "395.03", "city": "North Loriville", "zip": "48432"},
 {"id": "81e74ef5-e8e2-4d94-8ed9-04759531985d", "name": "Modular Blender",
  "price": "59.99", "city": "North Christopher", "zip": "55311"}]
Enter fullscreen mode Exit fullscreen mode

Output formats: JSON, CSV, or SQL INSERT statements, so generated data can go straight into a seed file.

Seeds make it deterministic

Every endpoint accepts a seed. Same seed, same data, byte for byte, seed 42 on the person generator gives you Margaret Johnson today, tomorrow, and in CI. That's the difference between a flaky fixture and a reliable one.

Locales are supported where they make sense (en_US, en_GB, de_DE, fr_FR, es_ES, ja_JP), and emails use safe, non-deliverable addresses.

Pricing

Every endpoint is available on every plan, including the free tier. All seven predefined generators and the custom schema endpoint are included from the start, with no Pro-only generators and no output format held back to force an upgrade. Upgrading simply gets you more requests per month: start free with 150 requests a month, move up to Pro at $4/month for 5,000, Ultra at $9/month for 20,000, or Mega at $24/month for 500,000. And since a single request can generate up to 1,000 records, even the free tier goes a long way. You pay for volume, not for access.

Try it: https://rapidapi.com/darksundarker/api/synthforge

What generator is missing for your use case? That's the v1.1 backlog, and it's being written by whoever asks.

Top comments (0)