About

Questions or comments about the project can be sent to the email address info@ipapi.is

Development on ipapi.is started in early 2021. Initially, the main purpose of the API was to check whether an IP address belongs to a hosting provider or not. Since the project gained slowly in popularity, the API was incrementally expanded, and the company, asn, and location objects were added to the API.

ipapi.is is a research project and acts primarily as a home for various interesting research ideas related to IP address data.

ipapi.is preferably uses public data to populate the database that is used for the API. For instance, IP geolocation intelligence is obtained from publicly downloadable WHOIS records and reverse IP lookups (dig -x 1.2.3.4). The is_vpn API flag is sourced from publicly available block lists such as github.com/X4BNet/lists_vpn. Another major data source is the awesome FireHol project, from which the API flags is_proxy (firehol_anonymous.netset) and is_abuser (firehol_abusers_1d.netset) are sourced.

The general motivation of this project is to prove that public data alone is sufficient to source a complete IP address API. A part of the data is freely available for download on the ipapi.is GitHub page. Another part of the ipapi.is research project is to publish the processes and mechanisms of obtaining the data used in the API.

But not only public data is used to populate the ipapi.is database as the following examples illustrate:

  • Hosting Detection Algorithm - A proprietary algorithm to detect previously unknown hosting/cloud IP ranges was developed. This makes ipapi.is the IP API with the best hosting detection quality available.
  • Network of HoneyPots - ipapi.is uses a vast network of honeypots distributed throughout the entire Internet to source the is_abuser API field.
  • Large Scraping/Crawling Pipeline - In order to keep the core data up to date and to enrich existing datasets, ipapi.is needs constant scraping and crawling. Millions of requests are fired from a distributed crawling architecture each week to keep the data up to date. For example, to derive the type field of the company and asn object, millions of company websites are crawled, and the crawled meta data is analyzed by current AI technology. Currently, the ChatGPT API is used to classify the company type based on the company's website meta data.

ipapi.is provides an extensive free plan, but a larger request volume (More than 30k monthly requests) is only available with purchased credits. Most similar projects either have a very limited free plan or don't include ASN and WHOIS information in their basic plan. With ipapi.is, you have access to all API output regardless of the subscription plan you choose. The only difference is in the request volume.

Up to 1,000 daily API requests are free. If you have a larger API volume, please consider subscribing to a billing plan or making a one-time credit purchase. Subscriptions are cancellable at any point in time. One-time credits never expire, and you can use them at your own pace.

API Strengths

Ipapi.is has several strengths:

  • IP Geolocation: By parsing WHOIS data and using open-source geolocation projects, an accurate and frequently updated geolocation database is compiled.
  • Accurate Hosting Detection: A complicated hosting detection algorithm uses public WHOIS data to detect thousands of hosting and cloud providers around the globe.
  • WHOIS Data: WHOIS data is the primary data source for IP address data. Since not all possible use cases for IP address data are known, ipapi.is provides raw and updated WHOIS records for all active ASN's and IP networks.
  • IP to ASN: The API provides accurate and frequently updated ASN information for all autonomous systems. If the ASN is active (at least one route assigned to it), WHOIS information is returned on top of it.
  • IP to Organization: Each IP address is owned (or allocated) by a company or organization. Knowing which organization is responsible for a certain IP range helps classify traffic originating from it in a better way. Organization data is also sourced from WHOIS records.

API Accuracy

There is a strong focus on data accuracy, but since the API consumes many third-party public IP address projects, it is impossible to guarantee error-free API output in all cases. Even WHOIS data, which otherwise is the most reliable data source for IP address data, may be erroneous at times.

In case an error is found in the data, please don't hesitate to contact ipapi.is via the email: info@ipapi.is.

Future Direction and Milestones

As of January 7, 2024

The data quality of ipapi.is is constantly improving. The following projects are planned for the future or currently in process:

  • Improve the organization coverage of ipapi.is data (Not yet started) - The task is to improve the organizational granularity of ipapi.is as it is seen in the company output object.
  • Systematically compare ipapi.is data quality with best competitors ipinfo.io, ipqualityscore.com, ipdata.co, ipapi.com (Not yet started) - The best kind of advertisement is to prove that ipapi.is data quality is superior in some aspects to the main competitors.
  • Derive IP geolocation intelligence from IPv4 Reverse DNS data (Not yet started) - Basically, the task is to replicate most of the findings from Microsoft Research paper "IP Geolocation through Reverse DNS"
  • Forward DNS Lookup of all known Domains (Not yet started) - The Internet consists of hundreds of millions domains from hundreds of TLDs. For many reasons, it is interesting to know which IP address hosts which domain name. Therefore, a complete forward DNS lookup of all known domain names is planned.

The following projects were already achieved and are completed:

  • Reverse DNS Lookup of the entire IPv4 Address Space (Started in Nov 2023, completed in Jan 2024) - Reverse DNS lookups for IP addresses can help to enrich the IP geolocation quality. Furthermore, reverse DNS data can be used to determine the connection type of an IP address. For that reason, a continuous reverse DNS lookup of the entire IPv4 space was conducted. The reverse DNS querying of 3.7 Billion (!) IPv4 addresses took roughly two months. The data will be used internally for ipapi.is, but the Reverse DNS database will also be sold directly on ipapi.is