> For the complete documentation index, see [llms.txt](https://docs.clearfeed.ai/clearfeed-help-center/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.clearfeed.ai/clearfeed-help-center/clearfeed-ai/ai-agents/bring-your-own-model.md).

# Bring Your Own Model

ClearFeed AI Agents use **large language models (LLMs)** to enhance your support workflows. By default, ClearFeed uses models from providers such as OpenAI and Google, chosen after rigorous internal evaluation to provide optimal accuracy and performance.

With the **Bring Your Own Model (BYOM)** feature, you can now use **your preferred LLM provider** for all AI tasks, ensuring your data stays within your approved providers.

### Supported Providers

ClearFeed supports any provider with an **OpenAI-compatible API**. Commonly supported providers include:

* OpenAI
* Anthropic
* Sambanova Systems
* Google Gemini

{% hint style="info" %}
BYOM configuration is currently not configurable through the ClearFeed UI. Contact ClearFeed Support to get this configured for your account.
{% endhint %}

### Configuring Models by Task Complexity

ClearFeed uses AI for a variety of tasks across your workflows. These tasks differ in how much reasoning they require, so balancing **latency, cost, and accuracy** is key. Examples of the kinds of tasks ClearFeed performs include:

* **Intent and urgency detection:** Classifying incoming messages to determine what they need and how urgent they are.
* **Auto-merging messages:** Grouping related messages into the same conversation or request.
* **Answer generation:** Producing accurate answers from your knowledge sources.
* **Planning and tool calling:** Deciding which steps to take and which external tools or integrations to invoke, and in what order.

With BYOM, ClearFeed works with you to establish a **mapping** from the models available on your provider to the relevant task categories above. This lets you assign the right model to each type of task — for example, using a faster, lower-cost model for simple classification while reserving a higher-capability model for answer generation. The mapping is tailored to your provider and requirements.

### Embedding Model Customization

In addition to task-specific models, you can configure the **embedding model** used for indexing your knowledge base into ClearFeed’s vector database. This affects:

* The **quality of search results** when retrieving relevant content.
* The **accuracy of references** used in AI-generated answers.

Choosing the right embedding model ensures that AI Agents provide precise, contextually relevant responses.

### Getting Started

To enable **Bring Your Own Model (BYOM)**, you’ll need the following:

* **API Key** from your chosen LLM provider.
* **Base URL** for API calls to the provider.
* **Model-to-Task mapping** specifying which models from your provider to use for each task category.

Discuss these details with the **ClearFeed Support team** to complete the setup.