Multi-model chat
In testingTalk to different models in one place and pick the profile that fits the task.
Ask anythingChat open for testing
What do you wantto get done today?
Pick the right model for each task, without leaving the conversation.
Create your account in under a minute and your question is waiting in the chat.
Available today and planned
Today, the chat. Next, a whole team of agents.
AI Workforce
Coming soonSpecialist agents coordinated by Core, with review before anything reaches you.
Missions
Coming soonScoped work with a defined deliverable and a human approval step.
Private AI
Coming soonModels that run in your environment, with no silent fallback to outside services.
How it works
Coming soonAsk. The team splits the work. You approve.
- 01
Ask
Type what you need, in your own words.
- 02
Delegate
Core picks the right specialists. Sentinel reviews.
- 03
Approve
Nothing ships without your OK.
Chat is live for testing. The agent team is planned.
How a task flows through the workforce
Authenticated access
The chat and platform areas require a signed-in account.
Human in the loop
Deliveries are meant to wait for your approval.
Controlled data flow
No external fallback without authorization.
FAQ
Is Orkestrion available today?
Partly. The chat is available for testing with an account. The AI Workforce, Missions, Model Hub catalog and public API are planned and not yet released.
Does my data leave my environment?
The Private profile is designed to keep content in your environment, with no fallback to an external service unless you authorize it. This is a design goal in development, not a certified guarantee.
Which models can I use?
The platform is built to work with open models served through Ollama and with OpenAI-compatible APIs. The list of specific models is still being defined.
Start with the chat.The rest arrives in stages.
Create your account in under a minute and your question is waiting in the chat.