Ollama: open models on your own computer, and how it compares
Ollama is one of the tools in the Codenation collection of local LLM tools. This article explains what it does, the jobs it fits, who it suits, and how it differs from the three other tools of that collection. Every price comes from the tool's own page, read on the date given at the end.
What Ollama is
Ollama downloads open models and runs them on your own computer. You use it from the command line, or through a local API that other programs call, so a script or an app can ask a model for an answer without sending anything to a hosted service.
It also sells access to larger models that run in Ollama's cloud, for when a model is too big for the machine at hand. Downloads exist for macOS, Windows and Linux, and the code is published on GitHub under the MIT licence.
Use cases
- Putting a local model behind a script, so text is processed on your own machine.
- Giving an editor plugin or an app a model to call through Ollama's local API.
- Serving local models to a terminal coding assistant: Aider names Ollama as a local runner it can work with.
- Moving to a larger hosted model through Ollama's cloud plans when the local hardware is not enough.
Who it is for
Developers who want a local model behind a script, an editor plugin or an app, and who are at ease in a terminal.
Someone who wants a chat window rather than a terminal will find LM Studio or Jan easier: both are desktop apps built around a conversation.
Price
Running models on your own hardware is free: the $0 Pay as you go plan has no subscription and no service fee. Cloud plans: Pro at $20 a month or $200 a year, with $60 of usage credits a month; Max at $100 a month; Team at $500 a month, in early access.
Read on ollama.com on 27 October 2026. Prices change: the official page is the reference.
Ollama and the alternatives
- LM Studio
- A desktop app to find, download and chat with open models, with a command-line tool and a local server that answers in the OpenAI API format. It suits people who want a graphical interface first. Running local models is free; hosted plans start at $20 a month.
- Jan
- A desktop chat app that runs open models offline and can also connect to hosted models from OpenAI, Anthropic, Mistral, Groq and others. Its code is open source under Apache 2.0, and its site lists no paid plan.
- Aider
- Not a model runner but a coding assistant for the terminal: it edits the files of a git repository and commits the change. It can use a local runner such as Ollama, so the two are often used together rather than one in place of the other.
Pick Ollama when a program, not a person, is the main user of the model: its local API is the point. Pick LM Studio or Jan when the main use is chatting with a model in a window. Add Aider on top of any of them when the job is editing code from the terminal.
Sources
Facts read on 27 October 2026. This article was written by Codenation. No one paid for it. A correction: contact@codenation.dev.