In the last few months, I did a Google search for the keyword Free Python Hosting. After doing some searching I got 3 free hosting which is pretty decent to try before switching to the premium (paid) version.
Before I discuss the free cloud hosting that I tried, this is my personal assessment by looking at several aspects that readers might use as references in choosing.
There are 3 Free Python Hosting or Python Cloud Hosting that I will review including:
Free Tier Python Hosting
|CPU||Low vCPU||0.2x – 0.5x vCPU||shared 1x vCPU|
|Support Cloud IDE||Yes||Yes||Yes|
|Database||MySQL||Replit Database||Heroku Postgresql|
Looking at the comparison table for free python hosting, each has almost the same average features in the Free Tier class. Free Tier allows us to try first, how the hosting service is, whether there are any problems during our app project.
As the name suggests: PythonAnywhere, we can deploy or edit “anywhere” anywhere as long as there is an internet connection using a laptop or smartphone. After logging in we can see a lot of information on the dashboard. For example – storage usage, CPU usage, recent files, used console.
- Support web apps: flask/django
- Deploy projects easily
- Suitable for testing small projects
- Schedule Task that allows us to perform certain processes every day
- File manager
- Free SSL for web apps
- Cloud IDE for mobile not working well
- Lack of supporting documentation
- The web interface is sometimes very slow and reloads suddenly
- Only 1 web apps project is allowed, and it runs for only 3 months and logs back in to activate again.
- There is no custom domain, so the web domain is accessed using a pythonanywhere address
Replit, personally, is my favorite cloud IDE for free python hosting. I tried to test live web apps 24/7 without any problems, indeed sometimes rebooting the system and loading start applications takes at least 4-5 minutes.
Setting the domain/subdomain which is very easy and free is one of the advantages of Replit. Where free tier users can use private domains/subdomains to be associated with the project.
Replit Universal Package Manager, we don’t need to be confused because in UPM we just need to do a package search and the package list appears after that we can install it with one click.
- Support web apps: flask/django
- Cloud IDE for mobile optimized
- Works on all web browsers
- Custom domain, can use our domain to be linked to the web apps project.
- Supports more than 50 programming languages
- Replit Universal Package Manager: one click install package
- File manager
- Free SSL of course when we launch web apps
- Server change suddenly
- Load modules/dependencies that take a long time to run the application
- Modules/dependencies are not installed automatically, will be reinstalled when the application is run
- Github repo link for premium users only
- Hidden repo only premium users
Actually Heroku doesn’t only support Python. There are several projects that can be deployed on Heroku such as: PHP, node.js, Ruby, Java, etc. In testing my application has been running for more than 2 months without any problems.
Free python hosting or Free Tier heroku provides 550 dyno hours, which means that our application can run non-stop for 550 hours (23 days if fulfilled), of course with the help of web monitoring. To be able to make our apps run for 30 days, we must verify by entering credit card information. But don’t worry, Heroku will not charge us if the use of addons or resources is still in the free tier category (pay as you go). Of course, with 550+450 dynos more than enough to run 1 web apps for 1 full month and still 256 dyno hours left if we count 31 days.
Heroku restarts the dynos I guess once every 24 hours. This is per dyno, not per application, random and does not coincide. If the dyno stops it will be restarted, and the 24 hour period reset (basically it’s a new dyno, new cycle).
- Support web apps
- Deploy via Github/CLI
- Additional 450dyno hours and custom domain, for verified users with credit cards.
- Very much Addon support.
- Adequate supporting documentation.
- Free SSL when launching web apps
- Database support, here heroku free tier can use Heroku Postgresql
- Ephemeral storage: means that the server can write to the storage and it doesn’t last long (temporarily), after the server restarts the storage returns to the same as when we deployed the project last time.
- Cloud IDE uses third-party, we have to install it first
- Schedule Task/Cron using dyno resources
- Premium tier is more expensive than other hosting.
Of the 3 free python hosting above, each has auto-sleep when there is no access/traffic for 30 minutes. As support, a web monitor is needed so that the application runs 24/7 non-stop like UpTimeRobot to ping our web application every 25 minutes. If asked which one is good? I will answer all good in the Free Tier class as a free python hosting provider.
All of the above free python hosting have their own advantages and disadvantages. So we can determine which free python hosting is suitable for our apps project. Personally I use Replit for testing projects and cloud IDEs, while production deploys to Heroku. Why not use PythonAnywhere, because PythonAnywhere is not yet optimal in mobile cloud IDE, I myself am working on a python project using a full smartphone, where the most optimal cloud IDE is Replit.
Tips & Advice
If you use free python hosting for our apps project, we need a little trick so that our project runs and functions properly. It does require a little more effort to make the project apps work if we use free python hosting, for example I use a free cloud json database, and Heroku Postgresql, and a private github repo, because it is not possible to save uploaded files on the Heroku server, and use third-party monitoring like UpTimeRobot to keep web apps running 24/7.
If using the 3 hosting above is considered suitable because of the features and performance it would be nice to upgrade the account to premium. If there is a credit card that still has a usage limit and has never been registered with AWS Free Tier, my advice is to switch to AWS Free Tier to try it, because they both have more capable features and specifications, like a private server if we can set it up and match it. our needs.
Calculate the premium tier price first, before switching to an upgrade. Because it could be more expensive, and also resources that are not really needed. Also compare the price of the python hosting with the VPS so that we can budget which one is appropriate.