Fake Data Generator

Q: Is the fake data real? Could it match a real person?

No. Every value is randomly assembled from token pools — any resemblance to a real person or business is coincidental. Emails use RFC 2606 reserved domains (example.com / .net / .org) that can never reach a real mailbox, and phone numbers are confined to the NANPA-reserved fictional block (XXX) 555-0100 through 555-0199, set aside so fake numbers never reach a real subscriber.

Q: Which output formats are supported?

CSV (RFC 4180, CRLF line endings, proper quoting), JSON (a pretty-printed array of objects with native types), SQL INSERT statements (with a configurable table name and escaped string literals), and TSV (tab-separated, for pasting into spreadsheets). Each download uses the correct MIME type and file extension.

Q: How do I reproduce the exact same dataset later?

Set a random seed (any text or number). The same seed plus the same settings always produces byte-identical output, which makes failing tests reproducible. Leave the seed blank for fresh random data on every click. The active seed is always shown so you can copy it for later.

Q: Is there a row limit?

Up to 100,000 rows. Everything is generated in your browser with no server, so the cap exists to keep the tab from freezing on huge datasets. Requests above the cap are clamped to 100,000 and you'll see a notice; values below 1 are clamped to 1.

Q: Does my data get uploaded anywhere?

No. Generation runs entirely client-side in your browser — nothing is sent to a server, which is exactly why this is safer than copying real customer or production data into a test environment.

Q: Are the generated addresses real and mailable?

No. Addresses combine plausible street, city, state and ZIP tokens but are NOT validated against USPS — a generated street + ZIP combination usually is not a real, deliverable address. Do not use this data for actual mailings or geocoding tests.

Generate realistic-but-fake sample data — names, emails, phones, addresses, companies, dates — as CSV, JSON, SQL or TSV. Privacy-safe, reproducible, built in your browser.

Rows & format

Number of rows

1–100,000, whole numbers. Generated entirely in your browser.

Output format

Columns (drag the ⠿ handle to reorder, or focus it and use arrow keys)

Field typeHeader / keyUniqueNull %

Pick the columns to emit, rename headers to match your schema, and reorder by dragging (or focus a ⠿ handle and use the arrow keys).

Date / datetime range

Start (ISO)

End (ISO)

Numeric range

Min

Max

Decimals

Decimals apply to decimal fields only (0–10). Integers use Min/Max as whole numbers.

Custom list values (one per line, or comma-separated)

A custom_list column picks a random value from this pool each row.

Options

Email domain

Reserved domains (RFC 2606) never reach a real mailbox.

Custom domain

Enabled when domain mode is set to custom.

Random seed (optional — for reproducible output)

Same seed + same settings = identical data. Leave blank for new random data on every generate.

Include header row (CSV / TSV)

Rows generated

—

View / copy full output

—

Rows generated—

Columns—

Format—

Estimated file size—

Fake Data Generator — guide & how to use

What it does

This free fake data generator builds realistic-but-fake sample data for seeding databases, testing forms, building demos, and filling spreadsheets — without ever touching real customer data. Pick how many rows you want, choose your columns (names, emails, phones, addresses, companies, dates, numbers, UUIDs and more), and export as CSV, JSON, SQL INSERT, or TSV. Everything runs in your browser; nothing is uploaded.

How to use it

Set the number of rows (1–100,000) and pick an output format.
Build your column set: add/remove columns, choose each field type, rename the header to match your schema, and toggle unique or a null % per column. Drag the handle to reorder.
Optionally set a date range, numeric range, custom list values, email-domain mode, and a random seed for reproducible output.
Click Generate — you'll see a live preview grid, per-column chips, and the full serialized output.
Copy the output or Download a correctly-named file (.csv, .json, .sql, or .tsv).

Worked example

You need to seed a customers table with 500 test rows that pass a UNIQUE constraint on email. Set rows = 500, format SQL INSERT, table name customers. Add columns id (sequential, always unique), full_name, email (toggle unique on), city, and signup_date (date, range = last 2 years). Set a seed like qa-2026 so the same dataset regenerates byte-for-byte if a test fails. Click Generate, then Download sample-data_500rows_….sql and run it against your test DB. Every email is a syntactically-valid address on a reserved domain that can never deliver — safe to seed anywhere.

Why it's privacy-safe (and the limits)

Emails never reach a real mailbox. The reserved-safe mode uses example.com / .net / .org, which RFC 2606 guarantees will never resolve to a real inbox. "Derive from company" and "custom domain" still build syntactically-valid addresses but a real custom domain could deliver — use it only for demos you control.
Phones are fictional by construction. All numbers fall in the NANPA-reserved block (NXX) 555-0100 through (NXX) 555-0199, set aside specifically so fictional numbers never reach a real subscriber.
UUIDs are RFC 4122 v4. Generated via crypto.randomUUID / crypto.getRandomValues where available — but treat them as test data, not real secrets.

Output format notes

CSV (RFC 4180): comma-delimited, CRLF line endings, fields containing commas, quotes, or newlines are double-quoted with embedded quotes doubled. ZIPs and IDs are emitted as text so leading zeros (07001) survive in spreadsheets. Some legacy Excel locales want a semicolon delimiter or a UTF-8 BOM — re-export elsewhere if your import target misbehaves.
JSON: a pretty-printed array of objects; integers/decimals are native numbers, booleans are true/false, blanks are null.
SQL: one INSERT INTO <table> (...) VALUES (...); per row, string literals single-quoted with ' escaped by doubling, blanks as NULL.
TSV: tab-delimited; tabs and newlines inside values are replaced with a space (TSV has no standard escaping).

Accuracy & honesty notes

This is FAKE data for dev / QA / demos only. Names, addresses, emails, and phone numbers are randomly assembled and are not real people or businesses; any resemblance is coincidental.
Addresses are not USPS-validated. A generated street + city + state + ZIP combination will usually not be a real, deliverable address. Do not use for actual mailing or geocoding.
Math.random() is not cryptographically secure. Non-sensitive fields use it; do not ship generated IDs or "tokens" as real secrets.
Seeded mode is deterministic by design and bypasses crypto — don't rely on seeded output for anything needing true randomness.
Unique columns can be exhausted. If a column's value space is smaller than the row count (e.g. unique boolean over 50 rows), the generator appends a numeric suffix and warns rather than looping forever.
100% private. All generation is client-side — nothing you configure or generate is ever sent to a server.

How to embed this tool on your website

No account, no coding. Copy the embed code and paste it where you want it to appear:

Embeddable in one line — copy the ready-made snippet from the box below.

Frequently asked questions

Is the fake data real? Could it match a real person?

No. Every value is randomly assembled from token pools — resemblance to a real person is coincidental. Emails use RFC 2606 reserved domains and phones use the 555-0100–555-0199 fictional block, so neither can reach a real person.

Which output formats are supported?

CSV (RFC 4180, CRLF), JSON (array of objects), SQL INSERT (configurable table name), and TSV. Each download uses the correct file extension and MIME type.

How do I reproduce the exact same dataset later?

Set a random seed. Same seed + same settings = byte-identical output, so failing tests are reproducible. Leave it blank for fresh data each click. The active seed is always shown so you can copy it.

Is there a row limit?

Up to 100,000 rows. The cap keeps generation entirely in-browser without freezing the tab; requests above it are clamped to 100,000 with a notice, and values below 1 clamp to 1.

Does my data get uploaded anywhere?

No. Generation runs entirely in your browser — which is exactly why it's safer than copying real customer or production data into a test environment.

Are the generated addresses real and mailable?

No. Addresses combine plausible tokens but are not USPS-validated; a generated street + ZIP usually is not a real, deliverable address. Don't use the data for mailings or geocoding.

Related tools

Put this fake data generator on your site

Free, one line, no account. Great for a developer docs page, an internal QA portal, or a tutorial.

<iframe src="https://snaptoolsuite.com/fake-data-generator/?embed=1"
  style="width:100%;max-width:620px;height:1200px;border:0;"
  title="Fake Data Generator" loading="lazy"></iframe>

2 Paste it on your page

3 It just works

Built locally in your browser. Nothing you configure or generate is ever sent, logged, or stored.