Hello, and welcome to this week’s changelog update! Let’s dive into what’s new:
-
Faster Koyeb Sandbox boot times
We’ve made significant improvements to the Koyeb Sandbox generation workflow, resulting in a measurable reduction in start time. This optimization makes it even faster to spin up isolated environments for development, testing, and automation. Check out the Koyeb Sandbox quickstart to get started with sanboxes.
-
Koyeb Sandboxes now support Nano Instances
When we launched our Koyeb Sandbox preview, you could deploy an Instance on any type (including GPUs) with the exception of Nano Instances. We’ve now added support for Nano Instances, expanding the range of workloads you can run in lightweight, ephemeral environments. This means all Instance types are now fully supported across Koyeb Sandboxes.
-
New tutorial - Generate Code Using OpenAI Codex in a Koyeb Sandbox Our new tutorial walks you through how to generate and execute code in a sandboxed environment using OpenAI Codex. It demonstrates how to take advantage of secure isolation while running dynamically generated code for development workflows. Check out the tutorial.
-
New one-click - Deploy PaddlePaddle FastDeploy
FastDeploy is an inference and deployment toolkit for large language models and visual language models based on PaddlePaddle. We’ve added it to our one-click deploy library, allowing you to launch it with the default baidu/ERNIE-4.5-0.3B-Paddle model or specify your own. Deploy FastDeploy on Koyeb.
