VectraGPSdocsGet a tracker
Apps & connectors

Connect ChatGPT to your fleet

Add Vectra as a custom connector in ChatGPT and ask it about your assets.

4 min read

ChatGPT reaches outside tools through custom connectors, which speak the same Model Context Protocol Vectra already serves. Same endpoint as every other assistant, same API key, same permissions.

Connecting

  1. 1
    Create an API key

    Settings → Apps → ChatGPT → Install. Leave the control toggle off for the first week.

  2. 2
    Turn on developer mode in ChatGPT

    Settings → Connectors → Advanced → Developer mode. Custom connectors live behind that switch.

  3. 3
    Create the connector

    Name it Vectra and paste the server URL below.

  4. 4
    Set authentication

    Pick the API-key option and paste your key. ChatGPT sends it as an Authorization: Bearer header, which Vectra accepts as well as X-API-Key.

  5. 5
    Ask it something

    Enable the connector in a chat, then try "which assets are moving right now?"

https://api.vectragps.com/mcp
The whole configuration. The endpoint is stateless, so there is no session to keep alive.
Developer mode moves around

OpenAI has shuffled where custom connectors live more than once. If the menu does not match, look for anything named connectors, apps or MCP servers in settings — the URL and key are the same wherever it ends up.

A connector is not a share link

The key carries your whole account at the permissions you gave it. To show one asset to one customer, use a share link instead — it expires and it shows nothing else.

NextConnect Cursor, Claude Code and other MCP clients
Something here wrong or missing? Tell us — these pages are maintained by the people who answer the phone.