Top Programming Languages Methodology 2025
Our best interactive programming languages are trying to solve the problem of estimating the popularity of a language by looking for proxy signals. We do this by building popularity measures from a variety of data sources which, in our view, are good proxys of active interest for each programming language. In total, we identify 64 Programming languages. We then lay each data source To create a global popularity indexexcluding some of the best scorers. Below, we describe the data sources we use to obtain the measurements and the weighting scheme that we use to produce the overall indices.
In popularity, we mean that we are trying to classify the languages that are in active use, including the maintenance activity of inherited systems. We examine three different aspects of popularity: languages in active use among the typical members of the IEEE and work software engineers (the spectrum classification)), the languages which are in demand by employers (the classification of “jobs”) and the languages which are in Zeitgeist (the “trend” classification).
We have evaluated the popularity of languages using the following sources for a total of seven measures (see below). We have collected information for all measures in July – August 2025. In the past, we strongly supported APIs to collect data from sources, but now the data is manually collected from the difficulty of following the API changes and endings, and Given that many names of the programming language (C ++, Scheme) have collided with common terms found in research documents and employment advertisements or have been difficult to analyze for a search engine. When a large number of research results have made it imprisoned to resolve ambiguities by examining all the results individually, we used a sample of each data source and determined the size of the relevant sample according to the estimation of the real average with 95%confidence. Not all data sources contain information for each programming language and we interpret this information as the programming language having “no blows” (that is to say not to be popular).
The results of each metric are standardized to produce a relative popularity score between 0 and 1. Then, individual measures are multiplied by a weight factor, combined and the renormalized result to produce an aggregated popularity score.By aggregating the measures, we hope to compensate for the statistical oddities that could distort the popularity score of a language in a particular source of data. The variation in weight factors allows us to create the different results for the spectrum, the work and the trendy rankings. We fully recognize that, although these weights are subjective, they are based on our understanding of sources and our prior coverage of software subjects. The variation in weight factors allows us to focus on different types of popularity and produce different rankings. We then combined each weighted data source for each program, then renounced the resulting frequency to produce an aggregated popularity score.
The best programming languages were originally created by data journalist Nick Diakopoulos. Our statistical methodology advisor is Hilary WETHING, although all of the data collection and calculation of the data are carried out by us. The rankings are calculated using R.
Google is the main search engine in the world, making it an ideal adjustment to estimate the popularity of language. We have measured the number of hits for each language by looking for the model, “X programming language” (with quotes) and manually recorded the number of results that have been returned by research. We took the measure in July 2025. We love this measure because it indicates the volume of online information resources on each programming language.
Battery overflow
Stack overflow is a popular site where programmers can ask questions about coding. We recorded the number of questions marked with each program during the last week before our research (August 2025). For the Mathematica / Wolfram language, we relied on the sister “Stack” for the Mathematica platform and counted the number of questions related to the programming posed during last week. These data were collected manually. This measure indicates which programming languages are currently in fashion.
IEEE maintains a digital library with millions of conference items and journals covering a wide range of scientific and engineering disciplines. We searched for newspaper articles, Magazine and Early Access which mention each of the languages of the “Programming” model for the beginning of 2025. For the search results which returned thousands of items, we identified the correct sample size for a 95% confidence interval (generally a little more than 300) and have drawn this number of articles. For each language that we have sample, we have identified the share of articles which use the programming language, then multiplied the total number of articles by this share to compensate for the total number of probable articles which references to a given programming language. We carried out this research in August 2025. This metric captures the prevalence of the various programming languages used and referenced in the Engineering Stock Exchange.
IEEE job site
We have measured the request of various programming languages in job offers on the IEEE job site. For the search results that have returned thousands of lists, we have identified the correct sample size for a 95% confidence interval (usually around 300 results) and have drawn this number of job lists to be examined manually. For each language that we have sample, we have identified the share of the lists which use the programming language, then multiplied the total number of job lists by this share to compensate for the total number of lists of probable jobs which refers to a given programming language. In addition, because some of the languages we follow could be ambiguous in raw text – such as Go, J, ADA and R – We have sought job offers with these words in the description of the position, then manually examined the results, again sampling the entries if the number of results were significant. The research was carried out in August 2025. We love the IEEE job site for its large number of non -American lists, making it an ideal to measure world popularity.
Career
We measured the request of various programming languages on the Careerbuilder employment site. We have looked for “developer” jobs offered in the United States because it is the most popular job for programmers. We sampered 400 job advertisements and examined them manually to identify the languages that employers mentioned in publications. The research was carried out in August 2024. We love the career site to identify the popularity of the programmer’s jobs among American companies
Github is a public repository for many open-source software projects focused on volunteers. We have used the GitHub list of its 50 best programming languages, filtering entries for things like Docker configuration scripts. The data cover the first quarter of 2025. This measure provides a strong indication in the languages that coders choose to work when they have a personal choice.
Library of Trinity College Dublin
The Trinity College Dublin library is one of the six legal deposits libraries in Ireland and the United Kingdom. A copy must be filed with the library of any printed material published or distributed in Ireland, and on request, any British publisher or distributor must also file a book. We have sought all the books published in the day to date which corresponded to the names of the programming languages and have verified the results of the false positives. The research was carried out in July 2025. We love this library collection because it represents a large and categorized sample of works, mainly in English.
Discord
Discord is a popular cat room platform where many programmers exchange information. We have counted the number of tags that correspond to each language. In the case of languages which could also be names of unscheduled subjects (many subjects linked to non-programming also have dedicated discord servers; for example, “Julia” could refer to the programming language or the Sesame Street puppet), the results were examined manual. Disboard was searched in August 2025. Disboard lists many public discord servers and many young coders use the site, contributing a different demography of coders.
