Skip to content
FlowtheraSI

Product

Private pilot

Flowthera Server

A secure, managed AI model server on your own machine. One install, an API your apps already understand, and updates you can roll back.

Request a pilotRead the concepts

What it does

Checks before it installs

Looks at your system, memory, disk, GPU and drivers first, and tells you plainly if something will not work.

Installs models and runtimes

Sets up the model runtime and downloads models from their publishers, checking each file against its published fingerprint.

Secure API with keys

Every request needs a key. Set limits per key and choose which models each key can use.

Health at a glance

See memory, GPU, disk and queue status, and benchmark models on your own hardware.

Signed updates, easy rollback

Updates are signed. If a new version fails its health check it rolls back automatically.

Private by default

Listens only on the machine itself until you choose otherwise. Logs record request details, never your prompts or answers.

Works with what you already use

Model runtimes

  • Ollama
  • llama.cpp
  • vLLM for GPU servers (planned)

Requirements at launch

  • Ubuntu 22.04 or 24.04 LTS on x86_64
  • CPU-only, or an NVIDIA GPU with a current driver
  • Bare metal, a VM or a cloud VM

Not yet supported: other Linux distributions, Windows Server, macOS as a server, ARM.

An API your tools already speak

The server offers OpenAI-compatible endpoints, so existing clients and libraries work by changing the address and key.

curl https://your-server.local/v1/chat/completions \
  -H "Authorization: Bearer fsk_your_key" \
  -H "Content-Type: application/json" \
  -d '{"model": "your-model", "messages": [{"role": "user", "content": "Hello"}]}'

Free and paid

Community, free

  • 1 machine
  • Installer, runtimes, secure API and model catalogue
  • 3 API keys
  • Security updates with rollback
  • Manual backup and restore
  • Docs and community support

Server licence

  • More machines
  • Any number of keys, with limits and model lists
  • Users and roles, alerts and usage
  • Scheduled encrypted backups
  • Email support

Pilot price: $99 per machine per month. See pricing.

Status: private pilot

We are installing it with a small number of pilot customers on their own hardware before a public release.

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.

Request a pilotJoin the waitlist