How it works
Your machines do the work
FlowtheraSI connects open AI models to the tools you use, on hardware you control. Here is where everything runs, and what leaves your machine.
The architecture
On your machines
Your apps
Flowthera Studio, Flowthera Mobile, or any tool that speaks the OpenAI-style API
Flowthera Server
API keys, limits, approvals, logs without prompt text
Model runtimes
Ollama and llama.cpp; external providers only if you switch them on
Your hardware
Laptop, workstation, Ubuntu VM or GPU server
FlowtheraSI cloud
flowthera.si website
Docs and downloads
Flowthera Account
Licences, billing, teams
Licence and update service
Signed licences and updates
No prompts, files or model output are stored here.
Four ways to install
A. Offline desktop
B. Your own Linux server
C. Team or private environment
D. Optional hybrid
What leaves your machine
| Data | By default | Sent to anyone? |
|---|---|---|
| Prompts, files and model output | Stay on your machines | Never |
| Crash reports and diagnostics | Off by default | Only if you switch them on |
| Requests to an external AI provider | Off by default | Only if you add your own key, and only to that provider |
| Licence checks and update downloads | Signed files fetched from our cloud | Account and licence details only |
A server you reach over the internet is private, but it is not offline. Read about security.
Run AI on infrastructure you control
We are working with a small group of pilot customers on their own hardware. Tell us what you want to achieve.