Browse categories
Explore
Fiverr Pro
English
$
USD
I will install ollama and open webui on self hosted vps as a private ai secure server
Mehmud
About this gig
If you want ChatGPT-style chat without sending company data to anyone, you run the model on your own machine. That is what this gig does.
I install Ollama with the Open WebUI interface on your server, on your own domain with SSL, and set up user accounts so your team logs in properly instead of sharing one link.
What you end up with:
A private chat interface at your own address, models pulled and ready to use, and a server that restarts cleanly and stays patched. On the Standard package I install NVIDIA drivers and CUDA if your box has a GPU, which is what makes the difference between fast answers and painful ones.
I will be honest with you about hardware before you spend money. Small models run acceptably on CPU. Anything in the 30B range needs a real GPU. Send me your server spec and I will tell you what will actually run on it rather than selling you a package that disappoints.
Advanced adds an API endpoint your own apps can call, and document upload so you can ask questions against your own files.
Message me with your specs first if you are not sure.
Get to know Mehmud
Mehmud
Server and Email Migration, n8n, Docker and Self Hosted AI Setup
- FromPakistan
- Member sinceJun 2016
- Last delivery2 years
Languages
Urdu, English
I have been running Linux and Windows servers for 9+ years - AWS, Google Cloud, DigitalOcean and plain VPS. Most of my work is migration: moving sites and data between servers, and moving business email from one provider to another without losing mail. I also set up self-hosted automation and AI stacks - n8n, Docker, Ollama with Open WebUI - for teams who want to own their infrastructure instead of monthly SaaS bills. And I clean up hacked servers and harden them so it does not happen again.
Tell me what you are running and where you want it. If it is not a job I should take, I will say so.
FAQ
What server do I need for this?
For 7B to 8B models on CPU, 4 vCPU and 16 GB RAM is a workable minimum. For anything larger, or for fast replies, you want an NVIDIA GPU with 16 GB or more of VRAM. Send me what you have and I will give you a straight answer before you order.
Which models can you install?
Anything in the Ollama library: Llama, Mistral, Qwen, Gemma, DeepSeek and the rest. Tell me what you want it for and I will suggest a size that fits your hardware.
Is my data really private?
Yes. The model runs on your machine and nothing is sent to an outside provider. There are no API keys to OpenAI or anyone else. That is the whole reason to self host rather than use a hosted service.
Can my whole team use it?
Yes. Open WebUI has proper user accounts and admin controls, so people log in with their own credentials instead of sharing one link. I set that up on the Standard and Advanced packages.
