Are you looking for a cost-effective way to experiment with cloud computing? Then, you should probably know about the Google Cloud Platform (GCP) Free Tier. GCP offers a Free Tier that provides you with a range of services free of charge. This blog will explore the GCP Free Tier, some use cases, and, more importantly, how to *stay* in the free tier.

What is the GCP Free Tier?

The GCP Free Tier is a set of services you can use without cost. GCP provides these services for free to help users explore the platform and to offer a risk-free way to experiment with cloud computing. The GCP Free Tier offers a wide range of services, including computing, storage, networking, and databases.

What do You Actually Get?

Straight-up $300 credit for every new account. More credits if you’re a business. You also get the following:

For computing power

• A month’s worth of compute on a small instance, with 30 GB of disk and 1GB of egress.
• 28 instance-hours on App Engine Standard
• Cloud Run for up to 2 million requests a month
• 2 million invocations for cloud functions
• 10GB of pub/sub messaging

For deploying apps (CI/CD)

• 120 build-minutes a day on cloud build
• One free cloud deploy pipeline
• Free monthly logging and metrics allotment
• 5 users, 50GB of storage and 50GB of egress on cloud source repositories

For Storage

• 5GB of standard storage on Cloud Storage (only US!), with 100GB of egress and 5,000 operations a month.
• 1TB of BigQuery queries a month, with 10GB of storage
• 1GB of a firestore database, with 50,000 reads, 20,000 wnites and 20,000 deletes a day per project.
• Secret Manager, with 6 active secret versions, and 10,000 access operations a month.

For AI

• 1,000 Vision AI calls.
• 5000 production units for natural language AI
• 500,000 translated characters a month
• 40 hours of training for video, with 5 hours for prediction
• 60 minutes of speech-to-text
• 1,000 minutes of video intelligence API invocations

An exhaustive list can be found here.

Note that if you’re a startup, you get a whole load of options to get free GCP goodies, too. So, what kind of things can we use the GCP free tier for?

Use Cases for the GCP Free Tier

GCP Free Tier is an excellent choice for developers, startups, and small businesses. Here are some use cases for the GCP Free Tier:

Building a website or webapp

If you have a business, chances are you need a website. You can host your website on GCP, or develop a web application all within the free tier.

• Compute Engine: Create a small virtual machine that can host your website.
• Cloud Storage: Store your website's files in Cloud Storage.
• Cloud Run: Hosting your web application.
• Firebase: Keep all of your user data in firebase databases.

Watchpoints

• If you need a domain, you’ll have to pay for that and a static IP if you’re hosting things on Compute Engine.
• If your traffic scales, you’ll need beefier VMs, meaning additional costs.
• If you go over your bandwidth limitations, Cloud CDN starts charging after 1TB of usage.

Developing a mobile app

Mobile applications need to use web services to run for logging, databases, and other things that make an application run.

• App Engine: Host your mobile app's backend services on App Engine.
• Cloud Storage: Store your app's data in Cloud Storage.
• Cloud Messaging: Send push notifications to your app's users.
• Firestore/Firebase: Useful for hosting everything you need to run a mobile app.

Watchpoints

• All of the above services are free for a certain amount. If you go over, you’ll have to pay.

Machine Learning

The GCP Free Tier also provides free services for machine learning. Here are some of the free services you can use for machine learning:

• AI Platform Notebooks: Provides a JupyterLab notebook environment with pre-configured TensorFlow and PyTorch environments. You can use this service to train machine learning models.
• AutoML Natural Language: This allows you to train custom machine learning models for natural language processing.
• Cloud Vision API: This allows you to integrate image analysis into your applications.
• Cloud Speech-to-Text: This allows you to convert spoken words into text.
• Cloud Translation: This allows you to translate text from one language to another.
• Cloud Video Intelligence: This allows you to extract metadata from videos.

Watchpoints

• With ML workloads, you really got to have a close eye on your usage. If you exceed the free limits, you will be charged for the excess usage.
• Be cautious of third-party services, which may have costs associated with them that are not covered by the free tier.

Archiving Data

You get a whole host of storage for free with GCP. Feel free to use it to archive the data that you may have lying around. Why not try to tie it in with some sort of backup system?

• Cloud Storage: 5GB of standard storage on Cloud Storage (only US!), with 100GB of egress and 5,000 operations a month.
• BigQuery: 1TB of BigQuery queries a month, with 10GB of storage.
• Firestore: 1GB of a Firestore database, with 50,000 reads, 20,000 writes, and 20,000 deletes a day per project.
• Secret Manager: Secret Manager, with 6 active secret versions and 10,000 access operations a month.

Watchpoints

• With storage, you got to keep an eye on how much you’re storing and the egress you’re using. GCP is nice and free when placing your data on it (ingress), but it costs to take it out (egress)!
• Don't forget to turn off services: If you are not using a service, turn it off to avoid unnecessary charges.

Centralising your data

BigQuery has a very generous free tier to play around with. It also comes with a nice sandbox for all the budding data engineers that want a playground to mess around with.

• Use BigLake (BigQuery + Google Cloud Storage) for your structured and unstructured data storage. You can query it for a real good amount (1TB of queries) without incurring any costs.

Watchpoints

• You know the script by now; monitor your usage!

Ways people have accidentally been charged while staying in the free tier

While the GCP Free Tier is designed to be risk-free, there have been cases where people have accidentally been charged while using only free tier services. Some common ways this can happen include:

• Exceeding free tier limits: If you exceed the limits of a free tier service, you may be charged for the excess usage.
• Using non-free tier services: If you accidentally use a non-free tier service, you may be charged for its usage.
• Not turning off services: If you leave a service running even when you're not using it, you may be charged for the time it is running.

To avoid any unexpected charges, it's important to keep track of your usage and be aware of the limitations of the free tier services.

Services Not Included in the GCP Free Tier

While the GCP Free Tier provides many services free of charge, there are some services that are not included in the free tier. These services have associated costs that can add up quickly if you're not careful. Some of these services include:

• Bigtable
• Cloud SQL for PostgreSQL
• Cloud SQL for SQL Server
• Dataflow
• Dataproc
• Datalab
• Dialogflow
• Kubernetes Engine (VMs it uses to run your stuff cost!)
• Memorystore
• Spanner
• Translation API
• Video Intelligence API

If you like a cloud horror story - here’s our favourite. It’s how Sudeep Chauhan burned $72k on Firebase due to a rogue exponential recursion function. It’s a brilliant read with some good lessons. Don’t be the person that lets their costs spiral out of control! And if you don’t know how to do that, we’re happy to lend you a helping hand 🤝.

Conclusion

The GCP Free Tier is an excellent option for developers, startups, and small businesses who want to try out cloud computing. It provides a range of services free of charge and allows you to experiment with cloud computing without much risk.

If you need help with any of the above, we have engineers that are chomping at the bit for problems to solve - so why not get in touch?