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

# Use Claude, ChatGPT, or Grok to Search Tendy via MCP

> Connect your preferred AI assistant to Tendy via the MCP integration and search, analyze, and track tenders without leaving the chat.

Tendy exposes a [Model Context Protocol (MCP)](https://modelcontextprotocol.io) server that lets external AI assistants query your procurement data in real time. Once connected, you can search tenders, retrieve full notices, and access your saved searches — all from inside the chat interface you already use every day, without opening Tendy in a separate tab.

## Supported assistants

The following AI assistants are verified and supported:

* **Claude** (Anthropic)
* **ChatGPT** (OpenAI)
* **Mistral Le Chat**
* **Grok** (xAI)

## What AI assistants can do via MCP

| Capability               | Description                                                                                  |
| ------------------------ | -------------------------------------------------------------------------------------------- |
| **Search tenders**       | Find tenders by keyword, procurement source, region, OKRB code, or estimated contract amount |
| **Get tender details**   | Retrieve the full notice for any tender, including dates, buyer info, and document links     |
| **List saved searches**  | View all the search templates you have saved in Tendy                                        |
| **Run a saved search**   | Execute any saved search template and get the latest matching results                        |
| **Access saved tenders** | Retrieve the tenders you have bookmarked for follow-up                                       |

## How to connect

There are two ways to authenticate the MCP server with your Tendy account.

### Option 1 — Personal Access Token (PAT)

Use a PAT for manual setup in any MCP-compatible client that is not yet listed as a verified connector.

1. Go to **Settings → Access Tokens** in Tendy.
2. Click **New token**, give it a descriptive name, and copy the token value.
3. In your MCP client, add a new server with:
   * **Endpoint:** `https://tendy.by/api/mcp`
   * **Auth header:** `Authorization: Bearer <your-token>`

### Option 2 — OAuth 2.1

Claude and ChatGPT connectors support a one-click OAuth 2.1 authorization flow. Open the connector page in the Integrations tab, click **Connect**, and authorize Tendy in the pop-up. No token copying required.

<Tip>
  For step-by-step instructions tailored to each assistant, see the [per-assistant setup guides](/integrations/mcp-connect) in the Integrations tab.
</Tip>

## Tendy AI vs. AI assistant integrations

<Note>
  **These are two different features.** AI assistant integrations (available on the **Basic plan** and higher) let you connect your own external AI tool — ChatGPT, Claude, and so on — to Tendy's procurement data. **Tendy AI** (available on the **Pro plan** and higher) is a separate built-in assistant that runs directly on tender documents inside Tendy itself. You can use both independently or together.
</Note>

## Plan requirement

AI assistant integrations via MCP are available on the **Basic plan** or higher. See [tendy.by/plans](/plans) for a full comparison.