Ollama
AdvancedOllama is a solid Gemini alternative. The local-AI standard for technical users. A terminal, one command, a model.
Information verified on July 22, 2026
Description
Ollama runs open source language models on your machine, from the command line or through its API. It has become the standard building block of local AI: one command downloads and runs a model. The project remains free and open source for local use; an optional paid cloud tier exists for large models, but nothing forces you into it.
My take
The tool for people at home in a terminal: the standard building block for running a model locally. I place it among the local options in chapter 13.
Key features
- Local execution in one command
- Local API
- macOS, Linux, Windows
- Library of open source models
- Optional cloud
Strengths
- + The de facto standard for local AI
- + API compatible with many apps
- + Very active open source project
Weaknesses
- − Command line: for users comfortable with a terminal
- − No built-in graphical interface
- − Large models need a lot of RAM
Recommended for
- Building local AI tools (API, scripts)
- Tinkerers who want to pick their models
Ollama vs Gemini: the comparison
| Criterion | Ollama | Gemini |
|---|---|---|
| Price | Free and open source locally. Optional cloud tier (Pro at $20/month). | Free, paid Google AI subscriptions |
| Targeted ads on your data | No | Conversations tied to your Google account, partly reviewed by humans |
| Source code | Open (auditable) | Proprietary |
| Migration difficulty | Moderate (3/5) | — |
Comparison
Ollama is the technical user's tool: scriptable, open, standard. For a terminal-free setup, GPT4All or LM Studio do the same job with an interface.
This topic is chapter 13 of my book
It answers these questions, among others:
- which assistant for which use, without ever trusting one player with everything?
- how do you run an AI at home, and on what machine?
- and how do you decide what you share, or not, with an online AI?
Forthcoming 22/10/2026 from Éditions Eyrolles
Frequently asked questions
How much does Ollama cost?
Free and open source locally. Optional cloud tier (Pro at $20/month).
How do you switch from Gemini to Ollama?
Estimated difficulty: Moderate (3/5). Install from the terminal, then one command per model. Easy if the command line doesn't scare you.
Who is Ollama for?
You are comfortable with a terminal and want the open, scriptable building block of local AI.
One spark a week
Every week in The Spark, I break down one tech story through the privacy lens. And what you can do about it, concretely.
Join 400+ readers. Unsubscribe in one click. Your data will never be shared.
Similar alternatives
GPT4All
BeginnerThe easiest local AI to install. Your questions never leave your computer.
LM Studio
IntermediateThe most polished interface for local AI, very much at home on Apple Silicon Macs.
Vibe (Mistral)
BeginnerVibe, the French chatbot by Mistral AI (formerly Le Chat). A solid everyday assistant, under GDPR rather than the Cloud Act.