The eight annual Python survey has been released. There are some interesting trends that can be seen in this survey and we will take a look at some of these in this blog. You can find the full survey here. The survey is the result of a fruitful collaboration between JetBrains the purveyors of the very widely used PyCharm Python IDE and the Python Software Foundation itself. In the most recent survey (where the data was actually collected in late 2024) over 30,000 developers, from a wide range of backgrounds, responded. This makes it the largest survey of its kind, ever!
Python usage
Perhaps the first thing to note about eh survey is the type of applications that developers are using Python for. As might well be expected the most common application for python is Data Analytics (49%). If you add in Data engineering at 33% then 82% of people are using Python as for Data oriented applications.
The next most common application type is Machine Learning which is often related to data analysis in some way as well. This application area has grown for just 34% to 42% in a year.
Interesting Web Development has also grown in the last year with the eighth survey indicating that 48% use Python to Web Development (whereas in the seventh survey it was 42%).
Experience with Python
The Python Survey asks all respondents about their level of experience with the language. Of those using Python 21% have been using it less than a year and 18% have only been using 1 to 2 years. That means 39% of python programmers have less than 2 years' experience with the language. When this is taken into account along with the number of years of professional coding experience you will see that 50% have less than 2 years programming experience (31% less than 1 year and 19% 1 to 2 years).
Find out about our Intro to Python training
Python versions
Back in the 80s and 90s a new version of a language might come out every few years. Fast forward to 2025 and developers now typically expect a new version of a programming language at least every year of not every six months. Python has a release cadence of a new version every October. This can be a problem for many organisations as it may take significant effort to accredit or validate a version of Python for use within the organisation; and with a yearly release they must plan to perform their validation every 12 months. This can have a negative effect on the availability of versions to developers within their organisation; with at least a new version taking some time to become generally available but also meaning that any existing applications may need to be retested or validated as well. This may well be reflected in the adoption of Python 3.X versions as represented by survey respondents’ data. 
According the eight Python survey, 35% of developers are still using Python 3.12. This version was released back in October 2023 and will be supported until 2028. However, there have been two versions since them; 3.13 (with 15% usage0 and 3.14 (with only 2% using it). These versions represent significant performance improvements in terms of performance and some language enhancements. In fact, the second most popular version is 3.11 with 21% of users and Python 3.10 with 15% of users are both even older and have significantly poorer performance than 3.14.
Use of AI
As with other surveys, such as the StackOverflow Developer Survey 2025, AI tools have crept into the Python survey. The Python survey illustrates that 82% of developers have used ChatGPT – which is a very significant number. It is well in excess of GitHub Copilot (39%), Google Gemini (23%) and Visual Studio IntelliCode (13%). Of course, it is not possible to say what the developers think of the support ChatGPT gives them only that it is the mostly used.
Find out about our AI-Augmented Python training.
Web frameworks
Web Development in Python is grown in prevalence for the eight survey, with FastAPI (38%), Django (35%) and Flask (34%) being the most popular. Although Django and Flask have changed little in popularity in the last few years, FastAPI has jumped significantly in popularity from 29% in 2023 to 38% in the latest survey. There are no obvious reasons for this but one can conjecture that FastAPI is very well designed for quick, fast, lightweight services that can be up and running very quickly. Given that many Python developers are new to the language and are working on data science style projects FastAPI may have a significant advantage here.
Find out about our training courses on FastAPI, Flask and Django
Machine Learning
Within the Machine Learning arena, the top four libraries are still SciKit-Learn (68%), PyTorch (66%), TensorFlow (49%) and SciPy (42%). These figures are mostly very similar to the 2023 figures with the exception of PyTorch which has grown from 60% to 66%.
Find out about our training courses on Machine Learning (Classic Algorithms), PyTorch and TensorFlow
Data Science
Data Science applications within the Python world are still dominated by the big two, that isa NumPy and Pandas. Of those involved in data exploration and processing a whopping 80% of developers use Pandas and 75% use NumPy. In contrast Polars is only used by 15% and Spark (aka PySpark) by 16%. As Polars positioned itself as a better, faster Pandas it seems not to be making that much of an in road into the Pandas world. Indeed, it appears that the newer, more recent versions of Pandas, have ensured that developers don't feel the need to switch.
Find out about our Python Data Science training programmes
Takeaways
Here are six takeaways from this blog:
If you are new to programming and / or new to python, then get on board, now is a great time to get started!
Be prepared for Python Developers to have limited software development and Python experience. This means that if you are writing documentation for them take that into consideration, equally it means that Python is viewed as an exciting area within which to work!
If you are going to use Python lean data processing and machine learning libraries as these are very widely used. In particular for data science learn Pandas and NumPy. For Machine learning get familiar with SciKit-Learn and / or PyTorch.
Update to the latest version of Python if you can.
If you want to use AI tools, then look at ChatGPT, it may or may not be the best but it is certainly the most widely used within the Python world.
If you are a web Developer, think about getting on board with FastAPI. If FastAPI doesn't meet your needs, then look at Django and Flask.
Would you like to know more?
If you found this article interesting you might be interested in our Python Training Courses.