← Catalog
sample data
Sample data downloads, free
Five sample files on this site, each free to download with no signup. Every dataset here ships a sample.json (schema plus the first rows) and a sample.csv (the same rows as CSV). They are plain static files: point curl at them, or download in a browser. Each card below states the exact columns and row count, and each sample link returned HTTP 200 on 2026-09-07.
Grab one in ten seconds
curl -s https://jayjex.github.io/data-vault/data/hud-fmr-2026/sample.json | jq '.columns, .records[0]'
curl -s -o hud-sample.csv https://jayjex.github.io/data-vault/data/hud-fmr-2026/sample.csv
The JSON shape is the same for every dataset: {slug, name, columns[], row_count, records[]}. Read columns to learn the schema, read records to see real values. The CSV twin has the identical rows with a header line, so it opens straight in Excel, Sheets, pandas, or any database importer.
What sample data is good for
Schema testing. Point your loader, migration, or validator at 10 to 22 real rows before committing to a 90,000-row import. Wrong types and missing values surface in seconds, and the columns listed below are the actual ones, so you can write your CREATE TABLE before downloading anything else.
Mockups and prototypes. A UI with real rows in it reads differently from one stuffed with foo and bar. Twenty NFL games with scores, spreads, and weather make a dashboard look finished; ten bounty listings with real reward amounts make a card layout testable.
Demo apps and fixtures. Sample files are small enough to commit to a repo as test fixtures. Deterministic, local, no API on the critical path: your CI loads a static file, not a live service.
Teaching. Explaining SQL joins or pandas groupby works better with an interesting table than with iris again. A 22-row game table with spreads and scores gives students something they already understand.
The five samples
sports
NFL games and betting lines, 22 rows × 46 columns
One row per game from the 1999-2026 pack. Columns include gameday, week, away_team, home_team, away_score, home_score, spread_line, total_line, away_moneyline, home_moneyline, rest days, starting QBs, coaches, and stadium weather (temp, wind, roof, surface). Wide enough to exercise any join or pivot you are building.
Sample: sample.json · sample.csv · Excel / Sheets import steps
Full pack: 7,548 games with spreads, totals, and moneylines, on Getly.
gov-data
HUD fair market rents FY2026, 22 rows × 10 columns
One row per ZIP code area. Columns: zip, hud_area_code, metro, area_name, state, and five rent columns fmr_0br through fmr_4br. Ten tidy columns, all numeric after the location fields, which makes it the friendliest of the five for a first import or a classroom demo.
Sample: sample.json · sample.csv
Full pack: 51,895 ZIP codes plus county and state tables, on Getly.
real-estate
Airbnb listings in 6 US cities, 20 rows × 12 columns
One row per listing. Columns: id, name, neighbourhood, latitude, longitude, room_type, price, minimum_nights, availability_365, number_of_reviews, reviews_per_month, last_review. The lat/lon pairs plot immediately, which makes it a good sample for map prototypes.
Sample: sample.json · sample.csv
Full pack: 90,169 listings across Austin, Nashville, Denver, NYC, Las Vegas, and San Diego, on Getly.
crypto
Superteam Earn bounty listings, 10 rows × 15 columns
One row per open listing. Columns: title, type, reward_amount, token, min_reward_ask, max_reward_ask, deadline, status, agent_access, is_pro, sponsor, comments, submissions, plus id and slug. Real rewards in USDC and other tokens, deadlines, and sponsor names.
Sample: sample.json · sample.csv
Full pack: 28 listings with full description HTML and 186+ reverse-engineered API routes, on Getly.
dev-tools
Web scraping script pack inventory, 10 rows × 2 columns
One row per script in the pack. Columns: file and pattern, the filename and what the script handles (pagination, infinite scroll, login sessions, price monitors, sitemap crawling, retries, and the rest). The smallest sample on the page, and a decent example of a minimal two-column inventory table.
Sample: sample.json · sample.csv
Full pack: ten Playwright scripts, Node 18+, on Getly.
Formats and licenses
Both files are static, served from GitHub Pages, no key, no rate limit beyond the host's. sample.json carries the schema; sample.csv carries the same rows with a header line. Licenses: the HUD sample is US government data, public domain; the NFL sample is CC BY 4.0 via nflverse; the Airbnb sample is CC BY 4.0 via Inside Airbnb. The license field in catalog.json states it per dataset if you need the exact string for attribution.