Local LLM tools
Jan: review, pricing, data and alternatives
Jan helps people run models on their computers and chat with local or remote models. It offers model downloads, an OpenAI API-compatible server, and connections between assistants and external tools. A free tier exists. Access and transparency score: 68/100, #4 of 6 in its category.
What is Jan?
Jan helps people run models on their computers and chat with local or remote models. It offers model downloads, an OpenAI API-compatible server, and connections between assistants and external tools.
According to its documentation, Jan lets you:
- run models on your computer
- chat with local or remote models
- download models from a catalog
- expose a server compatible with the OpenAI API
- connect assistants to external tools
We file it under Local LLM tools, the category of tools to run a model on your own machine, and it also overlaps with general AI chatbots and aI agent builders. The category lists 6 verified tools on this site.
The publisher is Menlo Research Pte Ltd.
Who is it for?
It is mostly useful to these profiles: developers and researchers. The need it covers is simple to state: run a model on your own machine.
Jan is a good candidate if:
- you want to try before paying
- you plan to connect it to your tools or automations through its API
- you must keep your data in Europe or on your own machine
- you want to read the code or host it yourself
Before deciding, run a trial on a real task and measure the time saved: it is the only criterion that really counts.
How much does Jan cost?
Pricing model: open source (public code). Basic use costs nothing.
To estimate what the same work would cost done directly with an AI model, use our API price calculator. Prices change often: always check the official site before subscribing.
Data and privacy
According to its privacy policy, data is processed on your device (local processing).
Local processing is the most favourable case: your files do not leave your machine for the features concerned.
Exporting your content: yes (details on the official site). Being able to get your data back makes switching tools easier later.
Three habits apply to every AI tool:
- turn off the use of your content for training if the option exists
- use a work account separate from your personal account
- review everything the tool produces before publishing or sending it
Platforms, languages and API
Jan runs on: Windows, macOS, Linux, and API.
Interface languages: Japanese.
A public API lets you call it from your own programs or an automation tool, and plug it into an AI agent.
The source code is public: janhqjanhq/jan.
How we scored it
Our 68/100 score does not judge output quality: it measures what can be publicly verified about access to the tool, its openness and how far the publisher can be trusted.
| Criterion | Points | What counts |
|---|---|---|
| Access | 15/30 | free tier, published price, French interface |
| Openness | 30/30 | API, data export, open code |
| Trust | 23/40 | identified publisher, declared hosting, age, presence in reference lists |
The full calculation is explained on our method page.
Comparison with alternatives
Jan against other tools in the local LLM tools category:
Verdict
Jan is a decent option; compare its price and terms with the category leaders.
Strengths
- Free tier to try it
- Public API
- Public source code
- Local processing available
- Available on many platforms
Alternatives
Categories and jobs
Learn more
Frequently asked questions
Is Jan free?
Yes, a free plan exists, with usage limits.
Where is Jan data stored?
According to the publisher, on your device (local processing).
Does Jan have an API?
Yes, a public API is documented.
What are the alternatives to Jan?
In the same category: LM Studio, PyGPT, Ollama, and GPT4All.
Jan or LM Studio: which one to choose?
Our score is 68/100 for Jan and 90/100 for LM Studio. Try both on the same task before paying.
Related articles
- Creating videos with AI: what works todayOctober 10, 2026
- Generating images with AI: tools, pricing, and limitationsOctober 9, 2026
- How to use NotebookLM: a complete guideOctober 8, 2026
- Free ChatGPT alternatives: what they actually offerOctober 7, 2026
Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score