Frequently Asked Questions


Wolfram Alpha LLC's goal is to cover all domains of systematic factual knowledge. The system is continually growing, but already includes by far the largest amount of computable knowledge ever assembled, spanning a huge range of areas. See the examples to get an indication of topics currently covered. Contact us to inquire about specific knowledge domains.

Wolfram|Alpha uses expertly reviewed curated data, trusted data providers and extensively validated internal algorithms. Unlike search-based systems, it does not rely on material dynamically retrieved from public web sources. The Wolfram|Alpha system is built on the Wolfram Language, one of the world's most respected software systems. Wolfram|Alpha uses extensive quality assurance on all components, and the system is continually updated to correct errors and omissions.

Not unless a contractual exception is negotiated; caching is generally inadvisable with Wolfram|Alpha, both because data is continually updated and because a surprisingly large fraction of results depends on user location (not only directly through geography, but also through unit, currency, linguistic and other conventions, as well as locally determined importance rankings of entities). The standard Wolfram|Alpha API contract does not allow caching.

Wolfram|Alpha APIs provide access to Wolfram|Alpha's large and continually growing collection of data. Contact us to suggest additional data for the system. Custom and enterprise Wolfram|Alpha services can use Wolfram|Alpha technology together with proprietary data.


Search engines, either interactively or through a programmatic API, give you links to pages that exist on the web. Wolfram|Alpha computes answers to specific questions using its built-in knowledgebase and algorithms.

Wikipedia gives you pages of narrative about topics. Wolfram|Alpha computes answers to specific questions you ask—just giving facts, not narratives. However, Wikipedia contains excellent "folk information" in many domains that may show up in results. Wolfram|Alpha uses such information, especially for linguistic disambiguation.

Not directly. Wolfram|Alpha has its own internal knowledgebase, with its own extensive internal semantics and ontology.

Traditional business intelligence systems rely on a strict schema for the data and encoded business logic for understanding relationships between datasets. Wolfram|Alpha uses a unique paradigm for modeling data, based on algorithmic computation, that allows for implicit relationships among the data and the emergence of new knowledge.


Wolfram|Alpha's data comes from its internal knowledgebase, curated by the Wolfram|Alpha team. Some of that data is derived from official public or private sources, but most of it is from more systematic primary sources. Source information is available for results, providing background sources and references.

Many kinds of data—such as financial and weather—stream in all the time. Other kinds of data are updated when they become available, on a daily, weekly, monthly, yearly or other basis.

No. The purpose of Wolfram|Alpha is to provide specific answers to questions, using both data and algorithms. The Wolfram|Alpha APIs are intended to provide fully processed answers, rather than raw data.

No. Almost every data result that Wolfram|Alpha gives is computed on the fly, so it would be impossible to download all of Wolfram|Alpha's data. Additionally, the data in Wolfram|Alpha's internal knowledgebase is updated continually; a download of the data would quickly become out of date.


Wolfram|Alpha introduces many new methods for understanding linguistic inputs. Mostly they're unlike traditional NLP, not least because Wolfram|Alpha has to deal with linguistic fragments rather than full grammatical sentences. Wolfram|Alpha does not rely on NLP methods to extract data from sources.

Yes. All Wolfram|Alpha APIs directly support specifying locations by latitude/longitude or IP address.

Its server technology is based on Apache web servers accessing clusters of Wolfram Engine (webMathematica) servers.

Because Wolfram|Alpha computes results in real time, it requires high-performance processors running in parallel and relatively little storage in its computational infrastructure. Web search is the opposite. As web search does not compute results, there is no need for fast processors, but there is a need for large amounts of distributed storage.

Wolfram|Alpha is designed from top to bottom for reliable operation. We have multiple distributed data centers. Within each data center, there are multiple independent Wolfram|Alpha server modules. Within each server module, there are multiple redundant Wolfram Engine instances. Each Wolfram Engine instance uses multiple redundant Wolfram Language kernels.


To access any Wolfram|Alpha API, you need two things: a current Wolfram ID and an AppID (API key). You may sign up for a Wolfram ID and an AppID online.

The AppID, also referred to as an Application ID or an API key, is a string that uniquely identifies your application to Wolfram|Alpha. The AppID must be sent with each Wolfram|Alpha API request. If you have multiple applications using the Wolfram|Alpha API, each must have its own AppID.

The Wolfram|Alpha APIs are standard REST APIs that can be used from any programming language. Requests are sent as a URL over HTTP, and responses may be XML or JSON. Explicit bindings are available for popular languages.

Not at this time. But please let us know if you are interested in using a SOAP interface rather than a REST interface.

Yes. The Wolfram|Alpha API is designed to operate with interactive web technologies such as AJAX, Flash and Silverlight.

Yes. The Full Results API allows you to select what subsets of full Wolfram|Alpha results you wish to receive, whereas other APIs are designed to choose only the most relevant results. Subsets often correspond to particular types of "pods" in output on the Wolfram|Alpha website.

The format of your output will vary depending on which API you're using, and with what options. When using the Full Results API, you can choose to receive output in various formats, such as plain text, images, sounds or copyable Wolfram Language input. Other Wolfram|Alpha APIs are designed to return specific formats; for example, the Simple API can only return an image, while the Short Answers API can only return plain text.

Yes. You can get images just like on the Wolfram|Alpha website. You can specify the formatting width to use; the APIs will automatically construct graphics and typeset material to that width.

It depends on the query and the type of result. Each API provides timeout specifications for optimizing response times. The Full Results API also allows specification of subsets of full Wolfram|Alpha output and provides options for obtaining results asynchronously.


Yes and no. Wolfram|Alpha APIs are available for free for non-commercial experimental use with a low monthly cap on queries. Once you're ready to take the next step, contact us for pricing options. The APIs access the complete Wolfram|Alpha system, with all its curated data and built-in algorithms, together with the Wolfram|Alpha supercomputer-class cloud infrastructure.

You can make up to 2,000 free API calls per month for non-commercial purposes. For commercial use and other options, the amount of requests you can make is subject to your pricing plan and the Wolfram|Alpha API Terms of Use.

Yes, upon entering into a commercial licensing agreement. Contact us to discuss a plan that suits your particular needs.

Yes. We have a full-service consulting organization that can handle projects small and large.

Yes. Wolfram|Alpha APIs can be used in any application, whether deployed to the web, desktop or mobile. The only system requirement is an active internet connection.

Yes. All you need is access to the internet, either by Wi-Fi or over a cellular network. The Wolfram|Alpha API is economical in the amount of data it sends.

General FAQ about Wolfram|Alpha ยป