AI & Machine Learning Jobs API - JobsIndex by Aspen Tech Labs

Access AI and machine learning job postings data through the JobsIndex API, built by Aspen Tech Labs.

JobsIndex aggregates 11M+ active jobs daily from 300K+ employer sources worldwide. Jobs are sourced directly from company career sites and ATS platforms, not scraped from job boards. That means cleaner data on AI jobs with direct application links where available.

This Actor serves two primary use cases: teams that want to publish AI and machine learning job listings on their platform, and teams that need structured job data for matching algorithms, AI model training, or workforce intelligence.

What You Can Use It For

• Power a job board, niche AI careers site, or job search feature with fresh, normalized AI and machine learning jobs
• Feed a matching platform or recommendation engine with structured AI jobs by role, skills, location, seniority, employer, and more
• Train or fine-tune AI models using real-world artificial intelligence job postings as source data
• Track AI and machine learning hiring demand across markets and geographies for workforce analytics or competitive intelligence

Built-In Search Preset

This Actor is fixed internally to category=Information Technology and sub_category=AI Specialist. Both fields are shown in the input example, and the runtime always enforces those values even if different values are submitted. Users can narrow the search by country, structured location, title, salary, company, posting age, employment type, or seniority.

Because the enforced category and sub-category take precedence over what in the current JobsIndex API, what is intentionally not exposed. Use the structured job filters below instead. where remains available for free-text location searches.

Quick Start

{
  "country": "United States",
  "size": 5
}

Input Parameters

Parameter	Description	Example
salary	Salary amount or range filter.	20h-30h
where	Free-text location search. Supports place names, boolean logic, and field-targeted search such as @country United States or @(city,metro_area) (Denver). Prefer structured location fields for exact filters.	Germany
country	Country filter.	United States
region	State, province, or region.	California
city	City filter.	New York
postal_code	Postal or ZIP code filter.	90210
metro_area	Metropolitan area filter.	Denver-Aurora-Centennial, CO
title	Job title filter.	Machine Learning Engineer
company_name	Company name filter.	IBM
company_domain	Company website domain.	ibm.com
category	Fixed preset. Runtime always uses Information Technology; changes are ignored.	Information Technology
sub_category	Fixed preset. Runtime always uses AI Specialist; changes are ignored.	AI Specialist
industry	Company industry filter.	Food Products
posted	Relative posting age or ISO date.	1w
employment_type	Employment type.	Full-Time
remote	Work arrangement filter. Supported values: remote, on-site, hybrid.	remote
seniority	Seniority level.	Senior
page	Page number, starting from 0.	0
size	Number of results per page. Default 5, min 1, max 100.	5

Salary Filter Format

Use:

NUMBER[UNIT]

or:

NUMBER[UNIT]-NUMBER[UNIT]

Supported units:

• h - per hour
• d - per day
• w - per week
• m - per month
• y - per year

Examples:

• 20h-30h - hourly rate between 20 and 30
• 35h - fixed hourly rate
• 20h-200000y - range from hourly to yearly comparison

Advanced Filters

where supports plain-text location search, boolean logic, and field-targeted search.

Examples:

where = United States
where = @country United States
where = @(country,region,city,metro_area,county,sub_city) (United States)
where = @(city,region,metro_area) (San Francisco OR California)

Structured Filter Precedence

When using where, avoid sending separate structured location filters for the same search unless you want those structured filters to take priority. Fields such as country, region, city, postal_code, and metro_area can override or narrow the free-text location query.

Use where as a free-text location search when structured location filters are not set.

Common location fields for where:

• country, region, city, metro_area, county, sub_city, postal_code

Location Filtering

For precise location matching, use structured filters exposed for this Actor:

• country
• region
• city
• postal_code
• metro_area

Use where only as a free-text location search when the structured location filters above are not set. If where is provided together with structured location filters, the structured filters take priority and where may be ignored or deprioritized.

Limits and Pagination

This Actor uses the JobsIndex jobs API. Results are paginated with page and size.

Parameter	Default	Min	Max
page	0	0	Depends on size and the item cap
size	5	1	100

Individual numeric bounds are checked before a run starts. Query-dependent pagination is checked after the run starts, as explained below.

One Run, One Page

Each run requests one page, not all available pages. size is the requested number of jobs for that page, with an absolute maximum of 100. Company filters can reduce the effective page size to 5 or 20; see Item Caps below.

For searches without company filters, the retrieval window is 1000 jobs: page * size must be less than 1000. The page number is zero-based.

size	Valid page range	First page outside the window
100	0-9	10
20	0-49	50
10	0-99	100
1	0-999	1000

The last page can contain fewer jobs than requested. For example, page=12, size=80 starts at job 960 and can return up to 40 jobs. Narrow the search with filters to retrieve a different result set beyond the window; the Actor does not change the requested page or paginate automatically.

The input schema checks page and size individually. It cannot pre-validate their combination: page=10, size=100 passes the individual bounds but is rejected after a run starts, so the run can still finish as FAILED.

Item Caps

The maximum number of retrievable jobs depends on the query shape:

Query shape	Max items	Effective max size	Over-size behavior
No company filter	1000	100	Validation error; not silently capped
Company filter plus another keyword or filter	20	20	Silent cap with capping metadata
Company filter only	5	5	Silent cap with capping metadata

A company filter means company_name or company_domain. Another filter means a value in fields such as title, category, sub_category, industry, posted, salary, employment_type, or seniority where available on this Actor.

When a company-filtered request asks for size greater than the effective cap, the API returns up to the capped number of results instead of erroring. The response metadata includes:

• size_requested - the original requested size
• size_capped_by - "company" for the 5-item cap or "company|keywords" for the 20-item cap

On uncapped responses, size_requested and size_capped_by are not present.

No-company-filter requests are not silently capped. If the request exceeds the 1000-item retrieval window, the API can return 400 Item limit reached. Maximum of 1000 items can be retrieved.

Free-Text Query Limits

These limits apply to each available user-entered field and are enforced before the Actor starts.

Input	Maximum characters	Maximum operator characters
where	500	20

The operator-character limit counts ?, +, pipe, and @ together within each input. Longer searches should be split into separate requests. Remaining backend syntax and safety checks can still reject an expression after the run starts.

The current JobsIndex API accepts up to 700 characters per effective query parameter. The smaller limits above, where applicable, and the 20-operator-character guard are Actor-side limits, not a claim that the API rejects every longer expression.

Field Validation

Apify validates the input schema before creating a run, including requests made directly through the API. A schema violation returns HTTP 400 with error.type = invalid-input; no run is created. API clients should handle this response when starting the Actor; there is no run ID to poll or run error file to fetch.

Input	Accepted values / limits
page	Integer 0-999; the usable maximum also depends on size and query shape.
size	Integer 1-100; company capping remains upstream.
company_domain	A domain name only, e.g. example.com; 3-100 raw input characters. No protocol, path, port, or internal spaces. The API trims surrounding ASCII whitespace.
posted	Recommended windows: 1d, 1w, 1m, 1y, or a real calendar date (YYYY, YYYY-MM, YYYY-MM-DD). Use 2-10 raw input characters. Surrounding ASCII whitespace is trimmed after the raw length check. Relative amounts use 1-4 digits without leading zeros.
salary	2-30 raw input characters. 1-9 digits plus h, d, w, m, or y; a range needs units on both ends, e.g. 20h-30h. Surrounding ASCII whitespace is trimmed. Decimals and currency prefixes are not accepted.
employment_type	2-50 raw input characters in total. Comma-separated values; 1-31 characters per item after trimming surrounding spaces. No empty items.
seniority	2-50 raw input characters in total. Comma-separated values; 1-31 characters per item after trimming surrounding spaces. No empty items.
remote	remote, on-site, hybrid, or an empty string for no filter.
Other text fields	Raw input character ranges: country (2-60), region (2-80), city (2-100), postal_code (3-15), metro_area (2-100), title (2-200), company_name (2-150), industry (2-100). Empty optional inputs remain allowed.

Empty optional text fields are allowed and are omitted from the upstream request. Raw input length means the decoded string including surrounding whitespace, before API trimming, not URL-encoded bytes. The schema checks preliminary formats, list structure, and lengths, not whether a matching location, employer, role, or job exists. The API retains its detailed character checks, including permitted Unicode characters in list values. A valid search with no matching jobs succeeds with an empty dataset.

Query-dependent pagination, combined filters, and other backend-only checks remain upstream or runtime checks. If they reject a request after a run starts, the run can still finish as FAILED. Technical errors are not reported as successful searches.

Output

Each run stores job records in the default Apify dataset. The dataset schema includes an Overview table for the most useful fields and a Raw data (all fields) table for the complete JobsIndex record.

Example dataset item:

{
  "id": "14242019290845606376",
  "title": "ML Engineer",
  "company_name": "Example AI Company",
  "category": "Information Technology",
  "sub_category": "AI Specialist",
  "country": "Canada",
  "region": "n/a",
  "city": "n/a",
  "salary_value": "178,000.00-228,000.00",
  "salary_currency": "USD",
  "salary_unit": "YEAR",
  "employment_type": "Full-Time",
  "remote": "remote",
  "posted": "2026-05-07",
  "url_apply": "https://example.com/apply"
}

Field availability depends on the source job posting.

Custom AI Jobs Data

For custom AI job data needs, bulk downloads, or tailored AI job feeds, contact us via our website or by email. You can also explore JobMarketPulse, our labor and job market intelligence platform.