I’ve been wondering for a while now if there’s a way to create an AI on Zima OS?
If so, which one do you recommend?
If not, why not?
How do they work in terms of
-
CPU resources? and
-
memory resources?
I don’t plan on using the graphics card for the AI.
Thank you.
Respectfully.
Best regards.
Not sure what you are planning to do with it. I have Zimaos running on a N100 Mini pc which has about 8gb ram. I have installed ollama with tinyllm, gemma3:4b and qwen3:4b. You would have to optimize it for your hardware but it is still slow. I have openwebui and hermes agent installed as well. my hermes is configure to use openrouter and my openwebui can use either of the llms, work with hemes agent or use openrouter directly. It all depends on what you are trying to achieve.
I use Claude AI, which I really like,
and I was wondering if there’s a way to use AI on my NAS?
Xeon E5 (12 cores/24 threads), 48 GB RAM, and over 30 TB of HDD storage.
I see that Ollama and Open WebUI are available, indeed.
I’m already using Ollama a (very little) bit locally on my PC (Gemma 4), but I definitely prefer Claude.
The problem is that it’s very limited and requires a paid subscription.
I’m not sure:
-
Is this a good thing?
-
Is it worth it?
Best regards.
Respectfully,
Dany.
Which one would you recommend?
I installed Open Web UI and Ollama CPU.
I added two engines:
Did I make a mistake?