Hosting Private Models And Tools

Run your own private models and tools without effort.

Introduction

TaskLift supports running private models and tools powered by TaskLift's scalable, secure and cost-efficient hosting platform. This page explains all the details related specifically to hosting but you'll find lots of related information on the following pages as well:

Server Lifecycle & Management

When the given private model or tool is first used, we start it by spinning up its server, bootstrapping the actual model or tool and connecting to it. This process usually takes around 10-20 seconds (may be longer for multiple/heavy models or multiple tools combined on a Computer).

If you make further queries to the same private model or tool within a next couple of minutes it will reply instantly and if you won't we will stop the server in order to cut the costs.

You may visit the Usage section from your user menu in order to:

  • see how many servers are running right now and how many servers have been started in the current billing period - along with the total hosting time and consumed credits
  • stop all the running servers instantly to cut costs or to start fresh - by clicking the Stop All Servers button that will appear if there's at least one running server

In case of any questions, issues or concerns related to TaskLift, don't hesistate to contact us.