Stratumus
Coming late 2026 Join early access

Your AI runs on your Mac. Use it from anywhere.

Stratumus connects local AI models on computers you own to one simple, private workspace. Install the Connector, pair your Mac, and use oMLX or Ollama without paying an inference provider for every request.

Join early access See how it works
  • macOS first
  • oMLX and Ollama
  • No paid inference API required
Illustrative preview of the Stratumus chat.

From download to private AI in three steps.

This is the installation experience currently being productized for early access.

  1. Install Stratumus Connector

    Open one macOS app. No Node.js, ports, or configuration files.

  2. Pair your Mac

    Approve the connection from stratumus.com.

  3. Start chatting

    Stratumus discovers oMLX, Ollama, and compatible installed models automatically.

Connector installedRuns in the background
Mac pairedApproved from stratumus.com
Model discoveredoMLX, Qwen3.5-9B-MLX-4bit
ReadyStart chatting

How a request travels.

A prompt entered at stratumus.com travels through an authenticated relay to your selected Mac. Your local model generates the response, and Stratumus streams it back to your browser.

Want the details of what each step can see? How your data moves.

  1. Your browser

    The Stratumus web app or PWA, wherever you are.

    Your device
  2. Encrypted connection

  3. Stratumus control plane

    Signs you in, keeps your conversations, and relays requests to your Node.

    Hosted by Stratumus
  4. Authenticated outbound relay

  5. Your Stratumus Connector

    Runs on your Mac and connects out to the control plane.

    Your Mac
  6. Loopback only

  7. oMLX or Ollama

    Runs the model you picked, on your hardware.

    Your Mac

Bring the local AI runtime you prefer.

Stratumus Nodes use a common provider interface, allowing the same chat experience to work with oMLX, Ollama, and future local engines.

Common provider interfaceThe same chat, whichever engine runs underneath

oMLX

Qwen3.5-9B-MLX-4bit
Supported in the current build

Ollama

llama3.2
Supported in the current build

More engines

Additional local runtimes
Planned

One workspace. Your choice of local AI.

Stratumus discovers supported providers and installed models on each connected Mac. Choose a primary Node, arrange fallbacks, and see the health of your private AI system at a glance.

Your NodesIn development
  • Home Mac oMLX · Qwen3.5-9B-MLX-4bit Healthy · Ready
    PRIMARYONLINE
  • MacBook Pro Ollama · llama3.2 Healthy · Ready
    FALLBACKONLINE
Preview of the Nodes screen. Primary and fallback management is in development.

The intelligence stays on your hardware.

Your selected model runs on your Stratumus Node, using hardware you own. Stratumus securely relays requests between the web app and your Node so you can use it from anywhere. No OpenAI, Anthropic, or other paid inference service is required.

Inference is local. Requests travel through a relay. We spell out both.

How your data moves

  • What the control plane handles

    Prompts and responses pass through the Stratumus control plane, currently hosted on Railway, on their way between your browser and your Node. Your browser connects over an encrypted connection, and your Connector connects outbound through an authenticated relay.

  • What is stored with conversations

    Conversations are managed by the Stratumus application, which is how your history persists between sessions.

  • What stays local

    Your models and inference. The selected model runs on your Node, and no third-party inference service is involved.

  • What changes with web search

    Web search is optional. When you turn it on, your search query goes to an external search service to fetch results.

  • What is planned

    Direct or end-to-end encrypted transport. Until then, requests are relayed through the control plane as described above.

Why run AI on hardware you own.

No bill for every request

Your model runs on a computer you already own, so you aren't paying an inference provider for each request. No OpenAI, Anthropic, or other paid inference service is required.

Your choice of local AI

Run the models and runtime you prefer, and keep the same chat experience when you change them.

Available when you're away

Your Node keeps working when you're not home. Use it from a browser or the Stratumus PWA, wherever you are.

What's proven, what's in progress, and what's next.

Validated in the current build

  • Secure authentication
  • Persistent conversations
  • Streaming responses
  • Optional web search
  • Local oMLX and Ollama inference
  • Multiple paired Nodes
  • Stratumus PWA
  • No required paid inference API
  • Authenticated outbound relay and reconnection

In development now

  • Downloadable macOS Connector
  • Browser-based pairing
  • Automatic provider and model discovery
  • Primary and fallback management
  • Rename and unpair controls
  • Clean installation without Terminal

Planned

  • Direct or end-to-end encrypted transport
  • Additional local engines beyond oMLX and Ollama
The Stratumus mark: a folded ribbon forming the letter S

Get early access.

Stratumus is starting on macOS with oMLX and Ollama. Leave your email and we'll let you know when early access opens.

  • macOS first
  • oMLX and Ollama
  • No paid inference API required