REVIEWS / AI CHATBOTS / OWNER INSIGHTS
🦉 WE READ 610 OWNER COMMENTS
Jan AI: what owners actually say
Owners praise Jan AI as the simplest local LLM tool they've used, but note it's still missing key features like image and document support
What owners complain about
- Missing core features FEW
Users report missing image upload and document handling capabilities, which they consider key functionalities for a local AI tool
- Model-specific issues FEW
One user had trouble with the tiny llama model — 'something is amiss' — and had to switch to downloading Mistral instead
- Speed vs cloud FEW
Local generation runs at around 200 words per minute on 16GB RAM, which the user notes is 'obviously not as fast as using a cloud provider'
- Unclear business model FEW
Users are asking whether it will remain completely open source and what the business model looks like, suggesting this isn't well communicated
What owners love
- Simplicity and ease of setup
Multiple users highlight the user-friendly interface and clean installation — one called it 'the simplest local AI I've installed so far,' another said 'no steep learning curves—just dive right in'
- Cross-platform support
Owners appreciate Windows, Mac, and Linux compatibility; one user confirmed it installs cleanly on Ubuntu
- Open source and active development
Users value the open-source nature, transparency, and an active project with 30+ contributors noted on Git stats
- Multiple model access
The model repository lets users access and switch between LLM models effortlessly — Mistral 7B, tiny llama, and GPT4 via API key all mentioned as working
- Offline capability
Owners praise it as an 'amazing offline solution' that runs locally without needing cloud connectivity
Surprising patterns
- 16GB of RAM is sufficient to run Mistral 7B locally at 200 words per minute — a concrete hardware benchmark not mentioned in marketing
- The tool supports both fully local models (Mistral, tiny llama) and cloud-based models (GPT4 via API key) in the same interface, making it a hybrid rather than purely offline
- Multiple commenters' primary reaction was simply asking about specs and GPT 3.5/4 comparisons, suggesting the product's positioning relative to mainstream LLMs isn't yet clear to new users
WHO SHOULD SKIP IT
Buyers who need image upload or document processing capabilities should wait, as these are currently missing according to owners.
Synthesised from 610 real owner comments across 3 platforms. Every point is grounded in the comments — no marketing, no AI guessing. How we do it →