New Eco Instances: the most affordable way to deploy apps globally
Today, we're making global deployment more accessible than ever with our new Eco Instance types, starting at $1.61 per month ($0.0000006 per second). We're also increasing the resources of the Free Instance type, giving you more to work with at no cost.
To make it even easier to get started on the platform, we’re introducing:
- The Eco Instance type: Half the price of the original Instances for the same amount of RAM. The eSmall Instance comes with 1GB of RAM and costs $5.36 per month or $0.000002 per second, effectively making it 50% cheaper than the standard small type ($10.72 per month).
- An improved Free Instance: We're doubling the RAM of the Free Instance and enabling deployments in Washington, D.C., in addition to Frankfurt, Germany. Each account can now deploy one Free Instance with 512MB RAM in the US or EU.
- Billing alerts: to prevent unexpected usage, we will notify you when you cross the $5, $20, $50, and $500 thresholds. If you're spending more, our customer success team will be in touch to set up custom notifications.
Designed for development and non-production environments, the new Eco Instances provide more RAM per core and are perfect for testing, prototyping, and experimenting before going to production. Here is the full lineup:
| Free | eNano | eMicro | eSmall | eMedium | eLarge | eXLarge | e2XLarge | |
|---|---|---|---|---|---|---|---|---|
| RAM | 512MB | 256 MB | 512 MB | 1 GB | 2 GB | 4 GB | 8 GB | 16 GB |
| eCPU | 0.1 | 0.1 | 0.25 | 0.5 | 1 | 2 | 4 | 8 |
| Price | Free | $1.61/mo $0.0000006/s | $2.68/mo $0.000001/s | $5.36/mo $0.000002/s | $10.71/mo $0.000004/s | $21.43/mo $0.000008/s | $42.85/mo $0.000016/s | $85.71/mo $0.000032/s |
These new Instance types are currently available for all of your Services running in Frankfurt, Germany and Washington, D.C. Backed by AMD EPYC Milan processors, they provide an outstanding price-performance ratio.
The Eco Instances benefit from all of the usual features of the platform including auto-healing, horizontal scaling, global load-balancing, native TLS with custom domains, a native VPC/mesh for each organization, built-in logs and metrics, continuous deployment, and more.
You can use the Eco Instances to deploy:
- APIs, full-stack web services, or workers
- More than 20 apps in one click at koyeb.com/deploy
- Using a GitHub repository, a Dockerfile, or a Docker registry
And you can pair them with our new Serverless Postgres databases.
Standard instances: more cores, available worldwide
In practice, nothing changes for existing workloads.
The current Instances are now called Standard. If you're new to Koyeb, the benefits of the Standard lineup include:
- 6 locations worldwide: The standard lineup is available for our 6 core locations worldwide. For now, the new Eco Instances are restricted to Frankfurt and Washington D.C., the locations where we have the most volume. We plan to add a location in Asia in the coming months for this new Eco lineup.
- More CPU: We allocate double the number of cores for Standard Instances compared to the Eco Instances.
- High-performance ephemeral NVMe storage: The Eco Instances have limited local ephemeral storage. If you need a lot of local SSD for caching, you'll want a standard Instance where you can get up to 400GB of high-performance NVMe storage.
The Standard Instances will also benefit from the latest generation of AMD EPYC CPUs, called Genoa, which provide 20% more performance per core.
Getting started
Eco Instances
The Eco Instances are available today from the control panel or the CLI.
Here is how to create an app called eco with an eSmall Instance (1GB of RAM) in Frankfurt and Washington, D.C.:
koyeb app init eco --docker koyeb/demo --regions fra,was --instance-type eco-small
Free Instance
If you have an existing Free Instance, your instance has already been upgraded to 512MB of RAM. You can also migrate your service to Washington, D.C. from the control panel or CLI.
Coming next for Instances: autoscaling, scale-to-zero, storage, and more
Our next features in line are autoscaling and scale-to-zero to provide a true serverless experience. We're also looking at supporting permanent NVMe storage for use cases that require high-performance local storage.
To get started, browse our selection of tutorials or deploy one of our One-Click Apps.
Join us on community.koyeb.com!