🍽️ Restaurant Health Inspection Scraper: 7 US Cities

Restaurant Health Inspection Scraper returns official health inspection results, individual violations and restaurant profiles from New York City, Chicago, Austin, King County (Seattle), New York State, San Francisco and Boston in one normalized dataset. Filter by name, ZIP, date, score and violation text, or monitor only new inspections. No login or API key. Pay per record: inspections $0.02, violations and profiles $0.01 each. Export to CSV, Excel or JSON.

RecordsData's Restaurant Health Inspection Scraper reads each jurisdiction's own open-data portal and maps seven schemas into one set of fields, paging each portal until maxItems is reached. In a cloud run on 2026-10-10 with all seven cities and inspectedSince 2026-09-01, it returned 150 rows (46 inspections, 59 violations, 45 profiles) in 11 seconds, and billing matched the delivered rows exactly.

🔎 What does the Restaurant Health Inspection Scraper do?

• Filter on the portal side by ZIP code, restaurant ID, date window, result text, score range, critical violations and violation keyword. Skipped rows cost nothing.
• Get one inspection-record per real inspection visit, one violation-record per cited violation, and one deduplicated restaurant-profile-record per restaurant.
• Compare cities with a normalized outcome and a scoreMeaning sentence, while result keeps the portal's own words.
• Turn on onlyNew to skip inspections delivered in earlier runs; skipped ones are not charged.
• Leave the name blank for the latest inspections; export to CSV, Excel, JSON or XML.

📋 What restaurant inspection data can you extract?

Each row uses the same 41 keys in a fixed order. The main ones:

Field	Type	Notes
recordType	string	inspection-record, violation-record or restaurant-profile-record
city, state	string	One of the seven jurisdictions and its state code
restaurantId, inspectionId	string	The portal's own ids; an id matches only in its own city
restaurantName, akaName	string	Business name; also-known-as name in Chicago
address, locality, zipCode	string	Address, borough or city, ZIP (none in San Francisco)
county, neighborhood	string	New York State county; San Francisco neighborhood
latitude, longitude	number	Not published by King County; zero coordinates in NYC are dropped
phone, cuisine	string	New York City only
facilityType, seatingRange, closedBusiness	string	Where the portal publishes them (King County for the last two)
inspectionDate, inspectionType	string	Future dates and dates before 2000 are never returned
result, resultDetail	string	Portal wording: grade, Pass, Fail, Satisfactory
outcome	string	Normalized: pass, pass-with-conditions, fail or closed
grade, score, scoreMeaning	string	NYC and King County grade; score where published; which direction is better
riskLevel	string	Chicago and King County
violationCount, criticalViolationCount	number or string	Counts; critical count in NYC, King County and New York State
violationCode, violationDescription	string	Violation records
violationComments, violationLevel, violationPoints, violationStatus	string	Chicago and Boston comments, King County level and points, Boston status
violationSummary	string	San Francisco only: one summary text per inspection
criticalFlag	string	Critical / Not Critical where the portal has it
source, scrapedAt	string	Link to the portal dataset; UTC time of the run

The original fields keep "N/A" where a portal has no value. Newer fields are left out when empty. Austin and San Francisco publish no per-violation rows.

🧾 Restaurant inspection scraper sample output

Real rows from a cloud run searching pizza (NYC, Chicago and Austin, maxItems 10). They predate the newest columns, so outcome, scoreMeaning and coordinates are not shown; see the table above.

{
  "recordType": "inspection-record",
  "city": "New York City",
  "restaurantId": "50041997",
  "restaurantName": "TOSKANA PIZZERIA",
  "inspectionDate": "2026-09-30T00:00:00.000",
  "result": "Z",
  "score": "57",
  "violationCount": 11,
  "scrapedAt": "2026-10-04T04:42:06.549Z"
}

{
  "recordType": "violation-record",
  "city": "New York City",
  "restaurantId": "50041997",
  "restaurantName": "TOSKANA PIZZERIA",
  "inspectionDate": "2026-09-30T00:00:00.000",
  "violationCode": "04L",
  "violationDescription": "Evidence of mice or live mice in establishment's food or non-food areas.",
  "criticalFlag": "Critical",
  "scrapedAt": "2026-10-04T04:42:07.039Z"
}

💰 How much does it cost to scrape restaurant health inspections?

Pay per event, only for records saved to your dataset.

Event	Price (free plan tier)	What it is
inspection-record	$0.02 ($0.0168 on Gold)	One inspection visit
violation-record	$0.01	One cited violation
restaurant-profile-record	$0.01	One unique restaurant, once per run

Worked examples at free plan prices. The 150-row cloud run (46 inspections, 59 violations, 45 profiles) cost 46 x $0.02 + 104 x $0.01 = $1.96. An 800-row run (221 inspections, 359 violations, 220 profiles) cost $4.42 + $5.79 = $10.21, about $12.76 per 1,000 records at that mix. 1,000 violation records alone cost $10.00. Failed fetches, empty searches, error rows and inspections skipped by onlyNew are never charged, and the run stops at your maximum charge limit. Paid plans get lower inspection prices; the Store page shows current ones.

🚀 How to scrape restaurant health inspections (3 steps)

1. Pick one or more cities.
2. Type a name such as pizza, add filters such as zipCodes or inspectedSince, and set maxItems.
3. Click Start, then download the dataset.

⚙️ Restaurant health inspection scraper input

Field	Type	Default	Description
cities	array	NYC, Chicago, Austin	nyc, chicago, austin, king-county, ny-state, san-francisco, boston
restaurantName	string	empty	Name to search for; blank returns the most recent inspections
zipCodes	array	empty	Only these 5-digit ZIP codes (San Francisco is skipped)
restaurantIds	array	empty	Only these restaurants, by restaurantId
inspectedSince, inspectedUntil	string	empty	Date window, YYYY-MM-DD
resultContains	string	empty	Text in the result, such as fail, pass or closed
minScore, maxScore	integer	empty	Score range; direction differs by city, see scoreMeaning
criticalOnly	boolean	false	Only critical violations (NYC, King County, New York State)
violationKeyword	string	empty	Text in the violation description, such as rodent
maxItems	integer	10	Max records across all cities and record types; free users are limited to 10
maxRowsPerCity	integer	0	Optional safety cap on source rows per city; 0 means no cap
onlyNew	boolean	false	Skip inspections already delivered for the same search
stateKey, resetSeen	string, boolean	empty, false	Name the monitor memory; forget it and start over

{ "cities": ["nyc", "king-county", "boston"], "inspectedSince": "2026-09-01", "criticalOnly": true, "onlyNew": true, "maxItems": 100 }

A city whose portal has no data for a filter is skipped and named in the run status instead of returning unfiltered rows; if every selected city is skipped the run fails with a message naming each, charging nothing. With criticalOnly or violationKeyword, violation rows and counts cover the matching violations only.

📦 Restaurant health inspection scraper output

The Overview view shows record type, city, restaurant, date, result, score and violation text. A search with no match finishes as SUCCEEDED with a status message and zero charges. If every city portal is unreachable, the run fails with an explicit message and nothing is charged.

Not delivered: Cincinnati, Montgomery County, and derived flags such as repeat offender, because a filtered search returns only part of a restaurant's history.

📊 Restaurant inspection scraper vs alternatives

Read on 2026-10-10 from each public Store page; "not stated" means their listing does not show it.

Dimension	This Actor	maydit	straightforward_hydra
Cities	7 incl. San Francisco, Boston, NY State	7 incl. Cincinnati, Boulder, Montgomery	5 incl. San Francisco
Filters	Name, ZIP, ID, dates, score, critical, keyword	Name, dates, result text	Name, ZIP, score, dates, keyword
Only-new mode	Yes, skipped rows are free	Not stated; dedupe left to you	Not stated; schedule with a date
Record types	Inspection, violation, profile rows	Flat rows with a granularity field	Flat rows, optional raw record
Price from	$0.02 inspection, $0.01 other	$2.40 per 1,000	$3.00 per 1,000

This Actor costs several times more per row than both. What the price buys is a deduplicated restaurant profile, a normalized outcome with score direction, and a monitor mode that skips seen inspections for free. Cincinnati, Boulder and Montgomery County are covered only by maydit. c0rrupt3d (NYC and Chicago, from $5.00 per 1,000) lists risk scores and repeat-offender flags that we do not compute. scrapemint ($10.00 per 1,000 rows) has a dedupe option too. If you need a flat, cheap feed, a rival fits better.

🧑‍💼 Restaurant inspection data use cases

Food safety and compliance teams, insurers, review and discovery apps, journalists and lead-generation teams (schedule onlyNew runs to catch restaurants that just failed an inspection) use this data.

🔌 Run via API, schedule and integrations

curl -X POST "https://api.apify.com/v2/acts/RecordsData~restaurant-health-inspection-aggregator/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"cities":["nyc","chicago","san-francisco"],"restaurantName":"pizza","inspectedSince":"2026-09-01","maxItems":10}'

Use the Apify client for Python or JavaScript, schedule daily runs, or connect the dataset to Zapier, Make, n8n or Google Sheets.

🤖 Restaurant health inspection scraper MCP server for AI agents (Claude, Cursor, VS Code)

Connect this actor to any MCP client through the Apify MCP server. Use this URL, which exposes only this actor as a tool:

https://mcp.apify.com?tools=RecordsData/restaurant-health-inspection-aggregator

Config for clients that take a JSON MCP config:

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com?tools=RecordsData/restaurant-health-inspection-aggregator"
    }
  }
}

Example agent prompt: "Use the restaurant health inspection scraper to find inspections since 2026-09-01 for restaurants named pizza in New York City and San Francisco, and list the critical violations with the inspection date. Run it again next week with onlyNew on to see just the new ones."

Choose this tool when an agent needs official restaurant health inspection results, violations or restaurant profiles for seven US jurisdictions. It takes the same input as the actor (cities, restaurantName, zipCodes, inspectedSince, minScore, criticalOnly, violationKeyword, onlyNew, maxItems and more); keep maxItems small for agent runs.

⚖️ Is it legal to scrape restaurant health inspection data?

The data comes from public open-data portals run by cities, counties and states, which publish it for public use. The Actor reads only that public data, needs no login and bypasses nothing. Operator, owner and inspector names are not delivered. Check each portal's terms and your local rules for how you use the data. This is not legal advice.

❓ Frequently asked questions

Which cities does the Restaurant Health Inspection Scraper cover?

The Restaurant Health Inspection Scraper covers seven jurisdictions, each read from its own official open-data portal: New York City, Chicago, Austin, King County (Seattle area), New York State outside NYC, San Francisco and Boston. Pick one or several in the cities field and the results come back in one normalized dataset. Cincinnati and Montgomery County are not covered yet.

Does the restaurant health inspection scraper include individual violations?

Yes for New York City, Chicago, King County, New York State and Boston: each cited violation is its own violation-record with code and description. San Francisco keeps its violations as one summary text per inspection, and Austin publishes only a score, so Austin returns inspection and profile records but no violation rows.

Why did I get 0 results from the Restaurant Health Inspection Scraper?

Your name or filters matched nothing in the selected cities, or a filter is not supported by a city and that city was skipped (the run status names it). Try a shorter name such as pizza, widen the date window or remove filters. A search with no match still ends as succeeded and the Restaurant Health Inspection Scraper charges nothing for it.

Why do restaurant inspection results look different per city?

Each portal publishes a different schema: NYC gives a letter grade plus a score where lower is better, Chicago gives Pass, Pass w/ Conditions or Fail, and Austin gives a score band. The Restaurant Health Inspection Scraper keeps the portal's own words in result, adds a normalized outcome (pass, pass-with-conditions, fail or closed) and a scoreMeaning sentence saying whether a higher or lower score is better.

Why are some restaurant inspection fields marked N/A?

The original fields (such as score, violation count and critical flag) keep the string N/A when a portal does not publish them, so existing pipelines do not break and you know it is a source limit. Newer fields, such as coordinates, phone and cuisine, are simply left out when empty. For example, Chicago publishes no numeric score and King County publishes no coordinates.

How is the restaurant profile billed in the Restaurant Health Inspection Scraper?

A restaurant-profile-record is charged once per unique restaurant in a run, not once per inspection, at $0.01. If the same restaurant appears in many inspection rows, the Restaurant Health Inspection Scraper still creates and bills only one profile record for it in that run.

Do I need an API key to scrape restaurant health inspections?

No. The Restaurant Health Inspection Scraper reads public open-data portals, so you need no login, no session and no API key. You only need an Apify account to run the actor, and free plan users get a 10-record preview per run, which is enough to check the output before you pay.

How fresh is the restaurant inspection data?

The data is as fresh as each portal, because the Restaurant Health Inspection Scraper reads them live on every run. In the 2026-10-10 cloud checks, inspections dated 2026-10-09 were returned. Schedule a daily run with onlyNew on to get only inspections you have not received before, at no charge for the ones skipped.

Why did my Restaurant Health Inspection Scraper run stop early?

The run reached your maxItems value, your maximum charge limit per run, or the optional maxRowsPerCity cap. A city with no new matching inspection for 45 seconds also stops, so an onlyNew run with nothing new ends quickly. Because each inspection is followed by its own violations, a small maxItems can be filled by one violation-heavy inspection.

🔗 Want more public-records data? Other RecordsData scrapers

• NHTSA Vehicle Safety Aggregator
• Healthcare Provider Payments Scraper
• Federal Contracts Spending Scraper

🆘 Support

Open an issue on the Actor's Issues tab or write to contact.punkrecordsdata@gmail.com. This is an independent tool, not affiliated with any of the listed cities, counties or states.

Last updated: 2026-10-11