# Adding and Managing LLMs

{% hint style="info" %}
Paid iMerit customers may, by contacting us [here](https://imerit.net/contact-us), request the integration of their own custom private LLM for fine-tuning, chain-of-prompting, real-time training, moderation, and many other possible workflows.
{% endhint %}

{% hint style="info" %}
LLMs may not be added or managed this way in private cloud and on-premise deployments of Ango Hub.

LLMs may still be part of projects in Ango Hub as plugins, but they may not be integrated with Ango Hub from the LLMs tab, and they may not be interacted with in LLM Chat-type assets.
{% endhint %}

Administrators can add large language models (LLMs) to their Ango Hub organizations. This allows users to interact with and prompt the LLM live through Ango Hub, on [LLM Chat-type assets](https://docs.imerit.net/labeling/labeling-editor-interface/llm-chat-labeling-editor). Messages in conversations with these LLMs, then, can be individually annotated with the [*Message*-type tool](https://docs.imerit.net/labeling/labeling-tools/tools/message).

## Adding an LLM

Navigate to the organization where you would like to add the LLM, and enter the *Organization* page.

Enter the *LLMs* tab. You will see a list of LLMs which have been added to the organization.

<div data-full-width="true"><figure><img src="https://3895963154-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FTcOUG6rfWxqGM0N4db2P%2Fuploads%2FEqA5Gmzw7Z9kyqL0Txng%2Fimage.png?alt=media&#x26;token=fd5b5d77-26c3-4b3f-8708-3edfe424d88d" alt=""><figcaption></figcaption></figure></div>

To add a new LLM, click on the *Add LLM* button. A dialog will appear.

<figure><img src="https://3895963154-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FTcOUG6rfWxqGM0N4db2P%2Fuploads%2FGokUB0tUpfXf2c1cJH6V%2Floreshot-20251031-144043%402x.png?alt=media&#x26;token=99a31323-a1df-437c-8714-70e88c996b52" alt="" width="563"><figcaption></figcaption></figure>

**Name:** The name you wish for your LLM to have on Ango Hub.

**Provider**: OpenRouter, OpenAI, or OpenAI-compatible.

**Model**: You may pick from this list any one of the models available through OpenRouter, including Llama, ChatGPT, Claude, DeepSeek, and many others.

**API Secret**: The API key from OpenRouter. Follow the instructions in the [section below](#obtaining-the-openrouter-api-key) for information on how to obtain it.

**Streamable**: If you enable this toggle, the answer from the LLM will come in streaming, rather than all at once. This allows the user to start reading the answer before it's fully formed, potentially improving TPT.

## How to obtain a OpenRouter API Key

From your OpenRouter dashboard, click on your profile picture, then on "Keys":

<figure><img src="https://3895963154-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FTcOUG6rfWxqGM0N4db2P%2Fuploads%2FothjrDCKbor941TqTzR7%2Fimage.png?alt=media&#x26;token=12304f79-468c-4a27-8eb5-e66df83f60e4" alt=""><figcaption></figcaption></figure>

Click on "Create Key". A new key will be created. Copy it and paste it into the "New LLM" dialog:

<figure><img src="https://3895963154-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FTcOUG6rfWxqGM0N4db2P%2Fuploads%2Fo1vABkugWYkTyRXNQrTg%2Fimage.png?alt=media&#x26;token=e7a743ff-793e-4c21-aeda-ee9502c1c44d" alt=""><figcaption></figcaption></figure>
