Chatbots & automation
Run Your Own Private ChatGPT-Style AI with Ollama and Open WebUI
A chat screen for you and your team that runs an open model on a server you own, so your prompts and documents never go to a third-party AI company.
Delivered in 1-2 working days
Fixed price · one-time · no hidden charges
or
Ask on WhatsApp before you buyA real person replies — usually within minutes during the day (IST).
How it works
- Pay ₹4,799Secure Razorpay checkout, one time. Then send the server details, domain and the model or use you have in mind.
- Installed and testedWe install Ollama and Open WebUI, add SSL, pull the model, lock down the admin and create the accounts.
- Backup proven and handed overWe test a backup restore, then send the guide and logins. You get 14 days of fixes.
The short answer
ZoopCoder installs Ollama and Open WebUI on your own Linux server for a fixed ₹4,799, ready in 1-2 working days. The price covers both tools in Docker Compose on a private network, one open-source model of your choice that fits your server’s memory, a chat address on your own domain with SSL, the admin account created by us with public sign-up switched off, up to 5 team accounts, a backup of the chat data with one tested restore, and 14 days of fixes. One honest rule shapes the setup: the speed of the answers depends on your hardware, and a server with no graphics card runs small models slowly, so we tell you what to expect for your server before you pay. The server and domain are billed separately.
What you get
- Ollama and Open WebUI installed with Docker Compose on your Ubuntu server, on a private Docker network so the Ollama port is never opened to the internet
- One open-source model of your choice pulled and tested, picked to fit your server’s RAM or graphics memory, plus a short note on what the model is and is not good at
- Your domain or subdomain pointed to the chat screen with free SSL that renews itself, HTTPS enforced, and the reverse proxy set so streaming replies work
- The first admin account created by us before the address is public, public sign-up switched off, and up to 5 team accounts created or invited
- A fixed secret key so people stay logged in after the container is recreated, and chat data kept on a named volume that survives updates
- Firewall on, root login and password login for SSH switched off
- A scheduled backup of the Open WebUI data folder to a location you choose, and one test restore
- A short written guide to adding another model and updating the containers, and 14 days of fixes for anything we configured
| Price | ₹4,799 (fixed) |
|---|---|
| Software | Ollama and Open WebUI, open-source (free) |
| Server | Your Ubuntu server, root access |
| Model | 1 open-source model of your choice |
| Accounts | Up to 5 |
| Domain and SSL | 1 domain or subdomain, free SSL |
| Support | 14 days of fixes after handover |
What we need from you
- An Ubuntu server with root access bought in your name — we tell you the size to buy for the model you want before you pay for one
- A domain or subdomain for the chat address, with DNS you can edit
- The model you want to run, or what the team will use it for so we can recommend one
- The email addresses of up to 5 team members
Not included
- The server fee, graphics-card server fee, domain fee and backup storage fee, billed by those providers to you
- Answer speed or quality at any level: both depend on your hardware and the model, and open models are not the same as the paid hosted ones
- Training or fine-tuning a model, or building a document search over your files — quoted separately
- Connecting the chat to your website, WhatsApp or other tools — see our website AI chatbot service
- More than one model, more than 5 accounts or a multi-server setup — quoted separately
About this service
Ollama is a free tool that downloads and runs open-source language models on your own machine, and Open WebUI is a free chat screen that sits on top of it and looks much like the hosted AI chat tools. Together they let a team ask questions, draft text and work with pasted documents without sending anything to an outside AI company, which is why businesses with client data like them. Both are simple to start and easy to get wrong on a public server. Ollama listens only on the server’s own address by default, and nothing protects it if you open that port, so we keep it on a private network. In Open WebUI the first account to sign up becomes the administrator, so we create it ourselves before the address is public. Without a fixed secret key everyone is logged out each time the container is recreated, and the chat data has to sit on a persistent volume or an update can wipe it. Models are held in memory for five minutes after use by default and a bigger context window uses more RAM, so we size the model to the server rather than to the brochure. If you want the AI on your website instead, see AI Chatbot for Website, and for the server itself see VPS Server Setup.
Questions before you buy
Is it free?
Both tools and the open models are free to run on your own server. You pay for the server, the domain and any storage for backups. If you later connect a paid AI service instead of a local model, that provider bills you separately.
Will it be as good as ChatGPT?
No promise of that. Open models you can run on an ordinary server are smaller and usually weaker than the best hosted ones. They are good for drafting, summarising and question answering on everyday work. Tell us the work and we will say honestly whether it is a fit before you pay.
Do I need a graphics card?
Not to start. A server with only a CPU can run small models, but answers arrive slowly. A server with a supported graphics card is much faster and costs more. We recommend a size for your model before you buy.
Is my data really private?
The prompts and chats stay on your server and are not sent to us or to an AI company. Anyone with access to the server can read them, so we limit logins to the accounts you approve.
Can you make it read my own documents?
Not in this service. Searching over your files needs extra setup and testing, so it is quoted separately. This service gives you the private chat and one working model.
Who owns the server and data?
You do. The server and domain are in your name and we only use the access you give us. You can remove our access once the 14 days of fixes are over.
Often ordered with this
© 2026 ZoopCoder