Workspaces
Each person gets an agent of their own: files, chats, memory, schedules, skills. Idle workspaces go to sleep and use no memory, and wake on the first request.
on your servers · under contract
EGGENT ENTERPRISE
All of Eggent, inside your perimeter.
We install the same stack that runs Eggent Cloud in your infrastructure: people's workspaces, the company console, the model gateway and file sync. Files, chats and the agents' memory stay with you.
WHAT WE INSTALL
The stack we run in our own cloud.
The company
The administrator's console: people, teams, limits, folders, messages, shared files and approvals, everything in Business.
The model gateway
One gateway for everyone. Provider keys live in it, not in the workspaces, and spend is counted per person and per team.
Files and Telegram
People's folders sync with their computers. The Telegram bot is yours, and the addresses are on your domain.
DATA AND MODELS
Your data, your models.
- Files, chats and the agents' memory are stored on your servers.
- Models run on your own keys: OpenAI or OpenRouter through the shared gateway, and in any workspace any provider, or a model of your own behind an OpenAI-compatible API, on your own hardware if you like.
- Keys in the gateway stay in the gateway: a workspace only gets its own pass to it.
- Model spend is counted per person and per team, with monthly limits.
- The Eggent core is open source under the MIT licence, so the agent's code can be read before it is installed.
- Your systems hand tasks to the agents over the API, and internal tools connect over MCP.
THE DIFFERENCE
The free install or Enterprise.
Open source · free
One workspace on your own server
- Installed with one command:
curl -fsSL https://get.eggent.ai | bash. - One agent with its projects, files, memory and schedules.
- Your own key or your own model.
- You update it yourself.
Enterprise · under contract
Eggent for the whole company
- A workspace for each person, a company with limits, shared files and approvals.
- A model gateway that counts spend per person and per team.
- Workspaces that sleep and wake, file sync, a Telegram bot of your own.
- We install, update and look after it.
WHAT YOU NEED
On your side.
- A Linux server with Docker, your own or in your cloud.
- A domain: workspaces live on its subdomains, and certificates are issued automatically.
- For scale: on our own server, 6 cores and 12 GB of memory keep about 30 workspaces running at once while the rest sleep.
- Model keys, or a model of your own.
HOW WE START
Four steps.
01
A conversation
How many people, where it should run, which models and internal systems it needs.
02
Installation
We install the stack on your server, connect the domain and add the first people.
03
Connections
Your models, your Telegram bot, your internal systems over MCP or the API.
04
Looking after it
Updates and support under a contract.
LET'S TALK
Tell us about your infrastructure.
How many people, where the data has to live, which models you use. We will come back with a setup and a price.