On-premise edition
An on-premise AI assistant for your whole team
An on-premise AI assistant runs on computers you own, inside your own network. SyncMind's on-premise edition keeps the models, the database, your chats and your files on those computers, so nothing leaves your office and it keeps working with no internet connection.
It is built for a team, not one person at one desk. Your team asks questions in shared chats, hands it documents, and gets finished work back: answers, Word documents, Excel sheets, slides and charts.
What running AI on-premise changes
Most AI assistants send each question to a company's servers to be answered. On-premise, the answer is worked out on your own machines instead. It is the strictest form of private AI for business. Three practical things follow:
- Your data stays put. Questions, documents and answers stay on your network. Nothing goes to a model provider. Here is how to check AI data privacy for business with any vendor.
- It works offline. No internet connection is needed to ask a question or work through a file.
- You are not paying per token. Usage is not metered, and there are no AI keys to manage, because the models are yours to run.
What SyncMind does on your own computers
Reads the documents your team already has
It reads PDF, Word, Excel, PowerPoint and CSV files, plus images and scanned pages, including tables and layouts. See how to chat with company documents.
Hands back finished work
It writes and runs Python to work through your files, then gives back real files anyone can open: Word documents, Excel sheets, PowerPoint decks and charts. Each run gets its own sandbox folder holding only the files it was given, and the code cannot open network connections.
Works with your team
Each person has personal chats, and teams share team chats. An administrator account and invite codes control who gets in. Working together like this is often called multiplayer AI. Deeper shared context for AI teams is listed for the API edition.
Shows every step
The Agent Log shows each step SyncMind takes as it takes it, so you can see how an answer was reached instead of taking it on trust.
Talks, on the computer itself
Voice runs on the computer too, so speaking to it does not send your voice anywhere. Phones can be paired by scanning a QR code on the same Wi-Fi.
Which models it runs, and on what hardware
Vendors often leave this out, so here it is. The on-premise edition runs models through Ollama, on your computers:
- Qwen 2.5 in 7B, 3B and 0.5B sizes, for answers
- Qwen 2.5 Coder, for writing code
- Granite Docling, for reading document layouts
You do not have to work out which one fits. SyncMind measures each computer it is installed on and downloads the models that computer can run. Computers on the network share the work, and a computer with too little power uses the team's AI over the network.
It installs with a desktop installer for Windows. If you are still deciding whether to run AI yourself, see what self-hosted AI for teams really takes.
On-premise edition or API edition?
SyncMind comes in two editions. On-premise gives up some convenience and features in exchange for keeping everything inside your building, so it is worth seeing the trade plainly.
| On-premise edition | API edition | |
|---|---|---|
| Where the AI runs | On your own computers | Hosted models, chosen by your plan |
| What leaves your network | Nothing | Questions, to the model provider; web searches |
| Works with no internet | Yes | No |
| Searches the web | No | Yes |
| Hardware | Uses each computer's own power | Any laptop, no GPU |
| Team knowledge, goals, roster | Not listed for this edition | Included |
| Pricing | Contact us; not metered by tokens | From $4 per seat a month |
Choose on-premise if everything must stay inside your building and you have computers to run it. Choose the API edition if you want it on any laptop, with web search and the full set of team features. The pricing page compares the two feature by feature.
Questions
Does my data leave the office?
With the on-premise edition, no. The models and the database run on your own computers, and it works without internet.
Do I need a GPU?
The on-premise edition measures each computer and downloads the models it can run. A computer with too little power uses the team's AI over the network.
What can it do with my files?
It reads PDFs, spreadsheets and images, including tables, layouts and scanned pages. It writes and runs Python to work with them, and hands back new files such as Word documents and Excel sheets.
Is the code it runs safe?
Each run gets its own sandbox folder with only the files it was given, and the code cannot open network connections.
Can I use it on my phone?
Yes. Pair a phone by scanning a QR code on the same Wi-Fi, then add it to the home screen. Chats you have opened stay readable when the computer is off.
How much does the on-premise edition cost?
Email contact@syncmind.in and we will talk it through with you. Usage is not metered by tokens, and you do not need AI keys.
Last updated