Deploy your own AI stack

Choose what you need, build a safe environment, then keep it running. Start small and verify each layer.

11 available steps · Open a guide for the details

01
Define the use case
Decide whether you need local inference, a chat interface or an application platform.
Read the guide →
Next: Choose local models
02
Choose local models
Understand Ollama's role before choosing hardware.
Read the guide →
Next: Size the server
03
Size the server
Match memory and compute to the intended workload.
Read the guide →
Next: Prepare Docker
04
Prepare Docker
Check volumes, networking and health before production use.
Read the guide →
Next: Install Ollama
05
Install Ollama
Install the runtime and verify it without exposing it publicly.
Read the guide →
Next: Add an interface
06
Add an interface
Pick the interface or platform that fits the task.
Read the guide →
Next: Check model setup
07
Check model setup
Understand the roles of model runtime and interface.
Read the guide →
Next: Connect a domain
08
Connect a domain
Add the appropriate DNS record and verify it.
Read the guide →
Next: Enable HTTPS
09
Enable HTTPS
Protect the public route before inviting users.
Read the guide →
Next: Secure access
10
Secure access
Review SSH, firewall and exposed services.
Read the guide →
Next: Back up and maintain
11
Back up and maintain
Know where data lives and rehearse a restore.
Read the guide →
Continue improving from real results

Define the use case