Overview
This help article will show you how to utilize iTwin IoT's MCP server with Gemini Enterprise.
Setting up the MCP Server for Gemini Enterprise
This step needs to be done by whomever is the admin for your organizations Gemini Enterprise account.
The first step is to set up the custom MCP Server connector for your organizations Gemini Enterprise account. A guide to set that up is shown here by Google. The URL for iTwin IoT's MCP server is https://itwiniotai-mcp.bentley.com/mcp and for Authorization it is https://itwiniotai-mcp.bentley.com/authorize
Client secret is an optional field
The client secret is optional.
Using the MCP Server in Gemini Enterprise
Once your Gemini Enterprise admin has done that you can enable the iTwin IoT MCP server connection by going to your Gemini chat, clicking the Connector icon and authorizing your iTwin IoT account via the iTwin IoT MCP connector.
Data handling and privacy
- What we collect. When Claude calls a tool, the connector logs the authenticated user identifier (your IMS
subclaim), the tool name, the duration of the call, and the outcome (success/error code). We do not log the contents of conversation messages or the raw parameter values you pass to tools. The data is kept for only 30 days and is used solely for monitoring, troubleshooting, and abuse prevention. - What we forward. Each tool call is authorised with your IMS access token. The iTwin IoT backend independently re-validates the token and enforces RBAC on every call; the connector cannot return data that your account is not entitled to.
- Storage. Audit logs are stored in Bentley's Azure tenant under the same governance as iTwin IoT itself.
- Third-party sharing. None beyond the iTwin IoT backend that you already use.
- Retention. Audit logs follow the default Bentley operational retention for monitoring data.
- Full privacy policy: see Bentley's global data privacy notice at https://www.bentley.com/legal/privacy-policy/.
Limits and quotas
To protect the iTwin IoT backend, the connector applies per-user sliding-window rate limits:
- 100 tool calls / 60 seconds per user (read-mostly tools).
- 30 tool calls / 60 seconds per user for destructive tools (
delete_*,cancel_*, mutating ML-job operations).
When you hit the limit, the connector returns a structured rate_limited error and the agent will wait and retry, or surface the limit to you to decide.