> ## Documentation Index
> Fetch the complete documentation index at: https://docs.powertokens.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# WorkBuddy Integration Guide

> Configure a custom model in WorkBuddy to connect to the powertokens platform via the OpenAI-compatible endpoint.

WorkBuddy is an AI office assistant client that supports connecting to third-party OpenAI-compatible endpoints via custom models. To integrate with the `powertokens platform`, fill in the platform endpoint URL, API Key, and model ID in WorkBuddy's model configuration, test the connection, and you're ready to go.

## Prerequisites

* A registered `powertokens platform` account with an API Key.
* WorkBuddy client installed (this guide is based on WorkBuddy 5.6.2).
* The model ID you want to use (e.g., `glm-5.3`, `MiniMax-M3`), which you can find in the platform model list.

## Configuration Steps

### 1. Open the model selector

In the WorkBuddy main interface, click the **⚡️ Fast** model selector at the bottom right to open the model dropdown menu.

At the bottom of the dropdown, click **Configure Custom Model**.

<img src="https://mintcdn.com/limeng/ES42zu3PD3dgWDUo/images/ecosystem-tools/workbuddy/step1-model-dropdown-en.png?fit=max&auto=format&n=ES42zu3PD3dgWDUo&q=85&s=055916435af796b5fdd78e89a8ed5fa3" alt="WorkBuddy model dropdown" width="1409" height="772" data-path="images/ecosystem-tools/workbuddy/step1-model-dropdown-en.png" />

### 2. Select custom provider

In the **Add Model** dialog, select **Custom** from the provider list.

<img src="https://mintcdn.com/limeng/ES42zu3PD3dgWDUo/images/ecosystem-tools/workbuddy/step2-select-custom-provider-en.png?fit=max&auto=format&n=ES42zu3PD3dgWDUo&q=85&s=aedd097a3146963151558014fab50be2" alt="Select custom provider" width="900" height="677" data-path="images/ecosystem-tools/workbuddy/step2-select-custom-provider-en.png" />

### 3. Fill in endpoint details and test connection

After selecting Custom, fill in the form with the following:

| Field | Value |
| - | - |
| **Endpoint URL** | `https://api.powertokens.ai/v1/chat/completions` |
| **API Key** | Your `powertokens platform` API Key |
| **Model Name** | A platform-supported model ID, e.g., `glm-5.3` |

After filling in the details, click **Test Connection**. Once the **Connection Successful** badge appears at the top, click **Save**.

<img src="https://mintcdn.com/limeng/ES42zu3PD3dgWDUo/images/ecosystem-tools/workbuddy/step3-configure-and-test-en.png?fit=max&auto=format&n=ES42zu3PD3dgWDUo&q=85&s=e9819a1de697c40c291bbf055de73737" alt="Fill in endpoint details and test connection" width="900" height="673" data-path="images/ecosystem-tools/workbuddy/step3-configure-and-test-en.png" />

### 4. Start using

After saving, select your newly added custom model from the model selector menu to start using it through the `powertokens platform` in WorkBuddy.

## Recommended Models

| Upstream Provider | Model ID |
| - | - |
| Zhipu | `glm-5.3`, `glm-5.2`, `glm-5-turbo` |
| MiniMax | `MiniMax-M3` |
| DeepSeek | `deepseek-v3-2-251201` |

## Limitations

* WorkBuddy custom models only support the OpenAI-compatible Chat Completions endpoint. Image generation, video generation, voice, and other task models are not supported.
* The **Tool Calling** option in advanced settings depends on model capabilities. Verify with real tasks.
* Do not expose real API Keys in documentation or screenshots.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.