<!-- Saycode On-Premise: run on your servers with an in-house LLM — markdown twin of https://saycode.ai/en/service/on-premise, for AI agents. -->
> Source: https://saycode.ai/en/service/on-premise
> Publisher: (주)버즈니 (Buzzni Inc.) · https://buzzni.com
> Summary: Run Saycode on-premise, on your servers or cloud, and connect an in-house LLM so AI requests stay on servers you run. We plan the location and setup with you.

ON-PREMISE

# On-premise installation and in-house LLMs

Saycode can also run on-premise, on your own servers or cloud. Connect an in-house LLM and AI requests stay on servers you run. We decide where and how it runs together, to fit your environment.

Contact salesStart

## Saycode Cloud vs. on-premise

Deployment options

Start on Saycode Cloud to use it right away, or on-premise if your data must stay in-house.

| Topic | Saycode Cloud | On-premise |
| --- | --- | --- |
| Where it runs | A cloud run by Saycode | Your servers or your cloud |
| AI models | Your AI subscriptions and APIs, plus in-house LLMs | In-house LLMs first; outside AI only when you approve it |
| Operations | Saycode runs and updates it | Runs inside your company; Saycode supports installation and updates |
| Getting started | Start right away | Install once the setup is agreed |

## AI models kept inside your company

In-house LLM

Once an admin registers your LLM server, employees pick it when they start a new chat.

Outside

Outside AI models · only when you approveApproved · encrypted

Inside the company

EmployeesNew chatSaycodeAssign · logIn-house LLMYour serversEmployee requests go to the in-house LLM inside your company. Outside AI models connect only through an approved, encrypted path when you allow them.

Steps

- Register serversAn admin adds your LLM server addresses. With several, conversations are spread across them.
- Assign by teamChoose who can use it, by group.
- Pick in a new chatEmployees choose the in-house LLM when they start a conversation.
- Usage on recordWho used how much shows in the admin console.

Works with LLM servers compatible with the Anthropic API.

## Frequently asked questions

### Can we install Saycode on our own servers?

Yes. Saycode can run on-premise, on your servers or your cloud. We plan the location, server setup and schedule with you during onboarding.

### Can we connect our in-house LLM?

Yes. An admin registers the LLM server address and chooses which teams can use it; employees then pick it in a new chat. The server needs to be compatible with the Anthropic API.

### Does data leave the company when we use an in-house LLM?

Requests sent to your in-house LLM go only to servers you run. If Saycode is installed on your servers too, conversations, code and results stay inside the company as well.

### Cloud or on-premise: which should we choose?

To try it quickly and grow from there, start on Saycode Cloud. If data can’t leave the company or you must use only in-house LLMs, on-premise is the fit. You can connect in-house LLMs in the cloud as well.

### How is on-premise priced?

We quote a one-time setup fee and a monthly operating fee, based on your company’s size and setup. Leave an inquiry and we’ll be in touch.

## Setup consultation for your environment

Leave an inquiry and we’ll review your environment and in-house LLM setup with you.

Contact salesStart
- Consultation
- Setup plan
- Installation
- Ongoing supportSee also[Enterprise](https://saycode.ai/en/service/enterprise)[AI cost savings](https://saycode.ai/en/service/ai-cost)[Compare](https://saycode.ai/en/service/comparison)
