Run AI projects
Train models, serve inference and process large datasets on dedicated GPUs.
Train models, run inference and render on dedicated GPUs. Buy credits with an local currency invoice and pay only while an instance runs.
Hourly billingCredits in birrChoose your GPUTrain models, serve inference and process large datasets on dedicated GPUs.
Pick a GPU and a software template. Stop it any time so you pay only while it runs.
Use the memory and parallel power you need for rendering, large models and batch jobs.
Your console lists the GPUs you can launch, their memory and hourly rates in birr, and your credit balance.
You buy credits in birr, and running instances use them by the minute at the hourly rate shown. The rate is locked when you launch. Each run has a one-minute minimum, and top-up invoices include VAT.
Yes. Start from a software template, then add the tools and libraries your project needs.
We stop the instance so billing ends, and email the workspace owner. It stays in your list, so you can top up and start it again.
Not on an existing instance. Launch a new one with the GPU you need, move your work across, then delete the old one.
No. Capacity depends on our GPU suppliers, and the largest GPUs are available on request.
With GPU partners outside Ethiopia, so data on these instances is stored outside the country. The location is shown before you launch.
Yes. Both sit in the same workspace and your VPS can call the GPU instance over a secure connection.
Talk through a website, application or agency use case before choosing a plan.