Arbeitsagentur Jobs Scraper (Germany API)

What does Arbeitsagentur Jobs Scraper do?

Arbeitsagentur Jobs Scraper collects job ads from the Jobsuche of the Bundesagentur für Arbeit, Germany's official and largest job board. You give it one or more search terms and, if you like, a location. It returns every matching ad as one clean JSON record with the job title, the employer, the work location (postcode, city, federal state), the publication date and a link to the ad on arbeitsagentur.de.

The actor uses the public Jobsuche API. No login, no browser and no proxy are needed, so runs are fast and stable.

Typical uses:

• follow the German job market for a profession or a region;
• build a list of employers that are hiring for a role;
• feed fresh job ads into a spreadsheet, a database or a job portal of your own.

How much does it cost?

You pay per job that the actor returns: $1.00 per 1,000 jobs ($0.001 per job).

• 1,000 jobs = $1.00
• 100 jobs = $0.10

Set Maximum number of jobs to cap the cost of a run. If you set a maximum charge for the run in Apify, the actor stops cleanly when that budget is reached and keeps everything it found until then.

How to use it

1. Open the actor and go to the input.
2. Enter one or more search terms, for example Softwareentwickler. German job titles give the best results.
3. Optionally set a location (city, postcode or region) and a radius in kilometres (at most 200, the limit of the API). Write the place name in German (München, not Munich; Köln, not Cologne) or use a postcode; an unknown place name fails that search with a clear message. Leave the location empty to search all of Germany.
4. Optionally limit the results to ads published within the last N days (0 to 100), and choose the offer type: jobs, apprenticeships and dual study, internships and trainee programmes, or self-employment.
5. Set Maximum number of jobs and start the run. Download the results as JSON, CSV or Excel, or get them through the Apify API.

Every search term is a separate search with the same filters. When the same ad matches two search terms, you get it (and pay for it) once.

Input and output example

Input:

{
    "searchQueries": ["Softwareentwickler", "Data Engineer"],
    "location": "Berlin",
    "radiusKm": 25,
    "publishedWithinDays": 30,
    "offerType": "1",
    "maxItems": 100
}

One item of the output (the values are an example):

{
    "source": "arbeitsagentur",
    "sourceJobId": "10000-1234567890-S",
    "title": "Softwareentwickler (m/w/d)",
    "company": "Beispiel Software GmbH",
    "locationText": "10115 Berlin",
    "city": "Berlin",
    "postalCode": "10115",
    "region": "BERLIN",
    "country": "DEUTSCHLAND",
    "employmentType": null,
    "postedAt": "2026-09-15",
    "description": null,
    "url": "https://www.arbeitsagentur.de/jobsuche/jobdetail/10000-1234567890-S",
    "inputUrl": null,
    "scrapedAt": "2026-10-05T12:00:00+00:00"
}

When an ad has more than one work location, locationText lists all of them and the other location fields describe the first one. region and country are written the way the source writes them.

Supported source

One source: the public Jobsuche API (version 6) of the Bundesagentur für Arbeit. It holds the ads that are published on the official job board of the agency. The actor uses only the open search endpoint and its publicly documented client id.

Limitations

• At most 10,000 results per search term. The API does not page past 10,000 results for one search. When a search has more, the actor keeps up to 10,000 per search term; duplicates the API itself returns are removed, so you may get slightly fewer. It logs a warning and goes on with the next term. Narrow the search with a location, a radius or a publication window, or split it into more specific terms.
• No job description in this version. The full ad text is only available through a detail endpoint that currently refuses requests. Open the url to read the ad.
• Private household employers get no name. When the employer name contains the word "Privathaushalt", "privater Haushalt" or "Privatperson", or is just "privat", company is null to protect private persons. Companies with "privat" in their name, such as "Privatklinik …", "Privat-Klinik … GmbH" or "… Privat Banking", keep their name. Employers that are listed with only a person's name are returned as listed.
• company is also null when the ad names no employer.
• employmentType and inputUrl are always null for this source. They exist so the output has the same shape as our other job scrapers.
• The data comes from the public Jobsuche API of the Bundesagentur für Arbeit. If the agency changes or closes the API, the actor stops with a clear error message until it is updated.

FAQ

Is it legal to use this data?

The actor reads publicly available job ads through a public API and does not return contact persons, e-mail addresses or phone numbers. You are responsible for how you use the data; check the terms of the Bundesagentur für Arbeit and the data protection rules that apply to you.

Why do I get fewer results than the source reports?

Duplicates between your search terms are removed, the maximum number of jobs or your budget may have been reached, or the search had more than 10,000 results (see Limitations). The log of the run shows how many matching ads the source reported for each search term.

Can I search in English?

You can, but most ads are written in German, so German job titles find far more results.

Can I run it every day?

Yes. Create a task with your input and add a schedule. A small Published within (days) value, such as 1, keeps each run to mostly new ads.