• Services
  • Pricing
  • Contact
Sign InBook Demo

Making construction more efficient through data.

Download on theApp StoreGet it on Google Play

© Copyright 2026 Bimefy BV. All Rights Reserved.
Company Registration Number: BE0800764890

Product
  • Features
  • Use Cases
  • Integrations
  • Pricing
  • Documentation
  • FAQ
Services
  • BIM-Coordination
  • BIM-Modelling
  • BIM-Management
  • BIM-Software
  • Customer Log In
Company
  • Bimefy
  • Blog
  • Contact
  • Careers
Supported by
  • imec.istart
  • Start it @KBC
  • VLAIO
  • AWS Activate
  • Microsoft for Startups
  • Hubs4Growth
  • POM Limburg
  • Smart Futures Lab
  • Flanders Investment & Trade
Legal
  • Terms of Service
  • Privacy Policy
  • Cookie Policy
  • Platform Uptime
  1. Integrations
  2. Model Context Protocol
Model Context Protocol logo

Developer & AI integration

Connect AI tools to Bimefy with MCP

The Model Context Protocol gives supported AI assistants a standard way to reach approved project data. Bimefy exposes a defined set of project tools, and you decide what a token can reach.

Request a demoView API documentation
Bimefy
  • Progress
  • Planning
  • Safety
Model Context Protocol logoModel Context Protocol
Answers grounded in this project's data

Connect AI assistants to Bimefy with MCP

Bimefy is a construction site intelligence platform. It connects BIM, construction planning, cameras, reality capture and site data so teams can compare what was planned with what is actually happening on site.

The Model Context Protocol (MCP) is an open standard for connecting AI assistants to external tools and data. Instead of each assistant needing a bespoke connector, a system exposes an MCP endpoint that describes the tools it offers, and a supported client connects to it.

Bimefy provides an MCP endpoint for its projects. Rather than opening the database to a model, it exposes a defined set of tools: specific, named operations such as retrieving current project progress, listing tasks, checking milestone status, summarising safety events or reading 4D view data. An assistant can call those tools and nothing else.

Access is controlled the same way as the rest of the Bimefy integration API. A project administrator issues a token, chooses the scopes it carries, and can revoke it. A token is bound to the project it was issued for, and the tools an assistant can see depend on the scopes granted. Requests are logged.

This is a deliberately narrow design. Bimefy does not hand an external AI system open access to project information, and the integration does not act on its own: an assistant answers questions using the tools you have approved, within the project you approved them for.

What the integration enables

What the MCP endpoint offers, and how it is contained.

A standard protocol

MCP is an open standard, so supported assistants connect without a bespoke integration being written for each one.

A defined set of tools

Bimefy exposes named operations covering project progress, tasks and milestones, safety and security events, site conditions, camera status and 4D data.

Scoped by token

Each token is bound to one project and carries only the scopes granted to it. The tools an assistant can see follow from those scopes.

Logged requests

Calls made with an integration token are logged, so it is possible to review what was requested.

Straightforward client setup

Bimefy provides the endpoint and a configuration snippet for supported MCP clients.

What the tools cover

Grouped by the kind of question they answer. Availability depends on the scopes on the token.

Progress and planning

Current project details, overall progress, tasks with filters for status and due date, milestone status including overdue milestones, and how model components changed over a chosen window.

Safety, security and site conditions

A current safety score with a daily trend, safety events including missing PPE, after-hours intrusion and person-detection events, grouped hazard detections, camera status, and IFC validation results.

4D views

The 4D views on a project, the catalogue of tasks within them, and the model elements those tasks relate to.

How the MCP integration works

Issue a token, point a client at the endpoint, ask questions.

  1. 1Create a token

    A project administrator issues a token in Bimefy's integration settings and selects the scopes it should carry.

  2. 2Configure the client

    Point a supported MCP client at the Bimefy endpoint using that token.

  3. 3Ask a question

    The assistant calls the Bimefy tools that its scopes allow.

  4. 4Get a grounded answer

    The answer is based on what those tools returned for that project, not on a general impression.

What teams do with it

Questions an assistant can answer from live project data.

Ask about current progress

Get the state of a project without opening a dashboard and reading it manually.

Summarise recent safety events

Pull safety events for a window and let the assistant summarise what came up.

Check milestones

Ask which milestones have passed their date and what the current status is.

Query 4D views

Ask about the tasks in a 4D view and the model elements they relate to.

Why MCP rather than a custom AI connector

Building a separate connector for every assistant does not scale, and each one becomes its own security review.

MCP moves that problem to a standard. Bimefy exposes one endpoint with a defined set of tools, and supported clients connect to it. The tools are explicit, so what an assistant can reach is a list rather than a guess.

It also keeps control where it belongs. Tokens are issued per project, carry only the scopes you grant, can be revoked, and are logged. Approved access, not open access.

Who is this for?

Roles that get the most out of this integration.

  • AI teams connecting assistants to real project data instead of exports.
  • Developers building tooling on top of construction project data.
  • Enterprise IT teams that need scoped, revocable, logged access.
  • Innovation managers piloting AI against live projects.

Frequently asked questions

What is MCP?

The Model Context Protocol (MCP) is an open standard for connecting AI assistants to external tools and data. A system exposes an endpoint describing the tools it offers, and supported clients connect to it without a bespoke integration.

What can an AI assistant do through the Bimefy MCP endpoint?

It can call a defined set of project tools covering progress, tasks and milestones, safety and security events, site conditions, camera status, IFC validation and 4D view data. It cannot do anything outside those tools.

Which AI clients can connect?

Any client that supports the Model Context Protocol. Bimefy provides the endpoint and a configuration snippet to set it up.

How is access controlled?

Access uses a token issued by a project administrator. A token is bound to a single project, carries only the scopes granted to it, can be revoked, and its requests are logged.

Does an external AI system get full access to our Bimefy data?

No. Access is limited to the tools Bimefy exposes and to the scopes on the token, within the project the token was issued for.

Can access be withdrawn?

Yes. Tokens are managed in the project's integration settings and can be rotated or revoked at any time.

Related integrations

Other systems Bimefy connects with

Bimefy API & custom integrations

Connect your own cameras, capture pipelines and sensor systems to Bimefy with scoped API tokens.

View integration

Autodesk

Bring Autodesk models, folders and file versions into Bimefy, and compare them with captured site reality.

View integration

Oracle Primavera

Connect construction schedules with BIM and captured site progress, and compare the programme with what the site shows.

View integration

Connect your construction data with site reality

See how Bimefy can connect your existing construction systems and turn site data into project intelligence.

Book a DemoStart Free
IBS
APK
PXL
Haex
Building Partners
Ibens Beyond
Strabag
Van Tornhaut
TechBIM
IBS
APK
PXL
Haex
Building Partners
Ibens Beyond
Strabag
Van Tornhaut
TechBIM

Trusted by innovation driven companies