How to Run AI Locally with Ollama (Step-by-Step)
Run powerful AI models on your own computer for free with Ollama. A step-by-step India-first guide with real commands for Windows, Mac and Linux.
ollama run llama3 — and you have a private AI assistant running offline on your own laptop, completely free. No API key, no subscription, no data sent to any company.Running AI locally means the model lives on your computer instead of a company's server. It works offline, keeps your data 100% private, and costs nothing per month. Ollama is the simplest free tool to do this — it handles downloading, running and chatting with open-source models in a single command.
What is Ollama and why use it?
- Free and offline: Run models like Llama 3, Mistral and Qwen with no monthly ₹ cost.
- Private: Your prompts never leave your machine — ideal for sensitive client or personal data.
- Simple: One command to download, one to chat. No coding required to start.
- Cross-platform: Works on Windows, macOS and Linux laptops.
How do I install Ollama on my computer?
Step 1: Download Ollama
Go to ollama.com and download the installer for your OS — Windows, macOS or Linux.
Step 2: Install it
Run the downloaded installer like any normal app. Ollama runs quietly in the background and adds the ollama command to your terminal.
Step 3: Verify the install
Open your terminal and check the version to confirm it works.
# Linux one-line install
curl -fsSL https://ollama.com/install.sh | sh
# Check Ollama is installed (all platforms)
ollama --versionHow do I download and run my first model?
Use ollama run followed by a model name. The first time, it downloads the model (a few GB); after that it runs instantly and offline.
# Download and chat with Llama 3 (general purpose)
ollama run llama3
# A smaller, faster model for low-RAM laptops
ollama run mistral
# Better Hindi / multilingual support
ollama run qwen
# Type your question, e.g. 'Explain photosynthesis in simple Hindi'
# Type /bye to exit the chatWhich Ollama commands should I know?
| Command | What it does |
|---|---|
| ollama run llama3 | Download (if needed) and start chatting with a model |
| ollama pull mistral | Download a model without starting chat |
| ollama list | Show all models you have downloaded |
| ollama rm qwen | Delete a model to free up disk space |
| ollama ps | Show models currently running in memory |
What hardware do I need, and what does it cost in India?
Pros
- A normal 8-16 GB RAM laptop runs 7B models well — no GPU needed
- Zero monthly cost; only electricity
- Works on existing hardware most students already own
Cons
- Big 70B models need a strong GPU or rented cloud GPU (₹40-₹120/hour)
- First download uses a few GB of data
- Older 4 GB RAM laptops struggle — stick to tiny models like Phi or Gemma 2B
Frequently asked questions
Is Ollama free to use in India?
Yes, completely free. Ollama is open-source and the models it runs are free to download. You only pay for electricity, and optionally a cloud GPU for very large models.
Does Ollama work without internet?
Yes. You need internet only once, to download a model. After that it runs fully offline — your prompts and data never leave your computer.
Can I run Ollama on a laptop without a graphics card?
Yes. 7B models like Llama 3, Mistral and Qwen run on CPU with 8-16 GB RAM. A GPU makes them faster but is not required to start.
Save this summary as an image or share it.
AICreatorHub Team
Hands-on AI practitioners covering tools, models and news for India.