H-1B LCA Scraper: Sponsors, Salaries and Approval Rates

This H-1B scraper reads the official labor condition applications that the US Department of Labor publishes for H-1B, H-1B1 and E-3 workers, and turns them into clean data you can filter, rank and chart. Look at single applications with the wage offered against the government prevailing wage, or skip straight to the answers: which companies sponsor the most workers, what they pay, how often they are certified, and what the market pays for a job.

Every wage is converted to a yearly amount, whether the employer typed it per hour, week or month. Typing mistakes in the source, such as a yearly salary written in the hourly field, are fixed and flagged, and impossible amounts are kept out of the statistics, so a single bad row cannot ruin an average. Employer rows are joined with the approvals and denials that USCIS publishes for each employer.

Job seekers, recruiters, immigration lawyers, compensation teams, researchers, journalists and investors use it to see who sponsors, for which jobs and at what pay.

What can you do with this H-1B scraper?

• Find the companies that sponsor a job title in a city or state, with their typical pay
• Benchmark salaries for a role: median, 10th and 90th percentile, by wage level
• Compare the wage offered with the prevailing wage and find offers far above or below it
• Rank sponsors by certified positions, with certification rate and approval rate
• Check an employer before you apply: filings per year, top jobs, wage levels, law firms, willful violator and H-1B dependent flags
• Follow decisions by date, for example every certification of the last month
• Look up approvals and denials per employer from the USCIS employer data hub
• Find small sponsors, new sponsors and staffing firms that place workers at other companies
• Work across several fiscal years in one run
• Export to CSV or Excel, or send new decisions to Slack and your own tools

What data can you extract from H-1B data?

Mode	One row is	Highlights
Filings	One labor condition application	Case number and outcome, visa class, received and decision dates, days to decision, job title, occupation code and name, full time flag, employment dates, worker positions, employment type, employer name, address, tax number, industry, H-1B dependent and willful violator flags, law firm, work city, county, state and ZIP, third party placement, offered wage and unit, yearly wage from and to, hourly equivalent, prevailing wage, wage level and source, offer versus prevailing wage in percent, wage correction and outlier flags
Sponsoring employers	One employer	Filings, certified, denied and withdrawn counts, certification rate, certified positions, median, average, 25th and 75th percentile, minimum and maximum yearly wage, median premium over prevailing wage, filings below prevailing wage, wage level mix, top job titles, occupations, work states and law firms, first and last decision, filings by fiscal year, USCIS initial and continuing approvals and denials and approval rate
Salary statistics	One occupation, occupation and state, or job title	Certified filings, median, average, 10th, 25th, 75th and 90th percentile wage, minimum and maximum, median prevailing wage, median premium, wage level mix, median wage per wage level, top employers with their median wage
USCIS approvals	One employer record of a year	Fiscal year, employer, city, state, ZIP, industry sector, tax number last four digits, initial and continuing approvals and denials, totals, approval rates, new sponsor flag

How to scrape H-1B data

1. Choose Filings, Sponsoring employers, Salary statistics or USCIS approvals.
2. Pick the fiscal years. The default is the latest year.
3. Add filters: employer names, job titles, states, cities, wage levels, minimum wage, outcomes.
4. Set Max Items and click Start.
5. Export the dataset as CSV, Excel or from the Apify console.

Each fiscal year is one big government file, so a run takes one to three minutes per year.

Input example

{
  "mode": "filings",
  "fiscalYears": [2026],
  "jobTitles": ["data scientist"],
  "worksiteStates": ["NY"],
  "caseStatuses": ["certified"],
  "sortBy": "annualWage",
  "maxItems": 100
}

Filing output example

{
  "caseNumber": "I-200-25308-369939",
  "caseStatus": "Certified",
  "visaClass": "H-1B",
  "employerName": "D. E. Shaw Research, Llc",
  "jobTitle": "Member of the Technical Staff",
  "socTitle": "Data Scientists",
  "worksiteCity": "New York",
  "worksiteState": "NY",
  "annualWageFromUsd": 731000,
  "prevailingWageAnnualUsd": 109491,
  "prevailingWageLevel": "II",
  "offeredVsPrevailingPct": 567.6,
  "decisionDate": "2025-11-14",
  "daysToDecision": 10,
  "lawFirm": "Fragomen, Del Rey, Bernsen & Loewy, Llp"
}

Employer output example

{
  "employerName": "Qualcomm Technologies, Inc.",
  "employerFein": "453157787",
  "filings": 813,
  "certifiedFilings": 811,
  "certificationRatePct": 100,
  "medianAnnualWageUsd": 152506,
  "p25AnnualWageUsd": 128898,
  "p75AnnualWageUsd": 179000,
  "isH1bDependent": true,
  "uscisFiscalYear": 2023,
  "uscisInitialApprovals": 30,
  "uscisApprovalRatePct": 96.1
}

Salary statistics output example

{
  "jobTitle": "Software Engineer",
  "certifiedFilings": 21876,
  "medianAnnualWageUsd": 135750,
  "p10AnnualWageUsd": 93490,
  "p90AnnualWageUsd": 229050,
  "medianWageByLevel": { "I": 101421, "II": 122667, "III": 180710, "IV": 226000 }
}

How much does it cost to scrape H-1B data?

You pay per result: $18 per 1,000 rows plus a tiny start fee. The free plan returns up to 10 rows per run. Employer and salary rows each summarize many applications, so a full list of 5,000 sponsors costs about ninety dollars and a salary table of 500 occupations about nine.

Tips for better results

• Yearly wages come from the wage offered, which is the minimum of the range. Bonus and stock pay are not part of it.
• The prevailing wage is the government benchmark for the job and place. Wage level I is entry level and IV is the most experienced.
• A rate of 100 percent certification is normal. Most applications are certified, and withdrawn cases are counted separately.
• An application is not a visa. One company can file many applications for one person, and approvals by USCIS are a later step. The USCIS rows show that step, and the latest year published is 2023.
• Employers are grouped by tax number when it exists and by name otherwise, so companies with several names show as one row.
• Staffing and consulting firms often place workers at other companies. The placed at third party flag marks it.
• Personal contact details of employer contacts and attorneys are not included on purpose. Only business names are.
• Add Only offers below the prevailing wage to see applications that need a closer look.
• For a weekly list of new decisions, schedule the Actor with the decision date from last week.

Who uses H-1B data?

• Job seekers and recruiters find sponsors and see real pay.
• Immigration lawyers research how an employer or an occupation fares.
• Compensation teams benchmark pay for engineers, analysts and nurses.
• Researchers and journalists study wages, dependence and denials.
• Investors follow hiring of tech and pharma companies.

Automate and connect

Schedule the Actor every month and send new sponsors to Slack, Google Sheets, Zapier, Make or n8n, or trigger a webhook when a run ends.

FAQ

Do I need an account or a login? No. The Actor works without any login.

Which years are available? Fiscal years 2020 to the current one, as published by the Department of Labor. The newest year grows every quarter.

Does it include denied and withdrawn applications? Yes. Use the outcome filter to keep only the ones you want.

Why is an hourly wage shown as a yearly amount? Every wage is converted to a year with 2,080 hours, 52 weeks or 12 months, so rows can be compared.

What does the wage corrected flag mean? The employer wrote a yearly amount in the hourly field. The Actor reads it as a yearly amount and marks the row.

Why are some employers missing approval rates? USCIS publishes the data by employer name and year, and some names do not match the labor department names.

Is this legal advice? No. It is public information about applications. Ask a lawyer before you make decisions about a visa.

Legal note

This Actor collects information that is publicly visible on the website. Check the source site's terms and your local rules before using the data.

Support

Missing a field or found a bug? Open an issue from the Actor page.