BIM-GPT: a Prompt-Based Virtual Assistant Framework for BIM Information Retrieval

BIM-GPT: a Prompt-Based Virtual Assistant Framework for BIM Information Retrieval
Do not index
Do not index
Blog URL
Efficient information retrieval (IR) from building information models (BIMs) poses significant challenges due to the necessity for deep BIM knowledge or extensive engineering efforts for automation. We introduce BIM-GPT, a prompt-based virtual assistant (VA) framework integrating BIM and generative pre-trained transformer (GPT) technologies to support NL-based IR. A prompt manager and dynamic template generate prompts for GPT models, enabling interpretation of NL queries, summarization of retrieved information, and answering BIM-related questions. In tests on a BIM IR dataset, our approach achieved 83.5% and 99.5% accuracy rates for classifying NL queries with no data and 2% data incorporated in prompts, respectively. Additionally, we validated the functionality of BIM-GPT through a VA prototype for a hospital building. This research contributes to the development of effective and versatile VAs for BIM IR in the construction industry, significantly enhancing BIM accessibility and reducing engineering efforts and training data requirements for processing NL queries.

Summary Notes

BIM-GPT: A Game Changer for Construction Information Retrieval

The construction sector is on a continuous quest to enhance the efficiency and precision of managing intricate projects. Central to this quest is Building Information Modeling (BIM), which digitally encapsulates the physical and functional traits of places.
However, the task of fetching specific information from BIM models can be daunting, especially for those not well-versed in BIM technology. This is where BIM-GPT enters the scene—a state-of-the-art virtual assistant that uses Generative Pre-trained Transformer (GPT) technology to streamline the retrieval of BIM information, making it easier and more user-friendly.

The Challenge at Hand

BIM technology encompasses detailed data from architectural designs to structural details, vital through a building's lifecycle. Yet, the sheer complexity and volume of data in BIM models can hinder information retrieval for non-experts.
Traditional BIM information retrieval methods demand a deep understanding of BIM systems and terminologies, posing a significant barrier to leveraging the full spectrum of BIM data.

Introducing BIM-GPT

The advent of AI, especially in natural language processing (NLP) with GPT models, provides a robust solution to these issues.
BIM-GPT utilizes this technology to allow for natural language queries in retrieving BIM information. This significantly lowers the entry barrier, broadening access to a wider audience.

BIM-GPT Architecture

The BIM-GPT system is built on three main modules:
  • User Interface (UI) Module: A user-friendly web platform for inputting natural language queries and visualizing BIM data.
  • Natural Language Processing (NLP) Module: The core of BIM-GPT, it uses GPT models to understand queries, find relevant information, and craft coherent responses.
  • Data Management (DM) Module: Handles the retrieval and management of BIM data based on the NLP module's processed queries.

Evaluating the Framework

Testing on an enhanced BIM information retrieval dataset showed BIM-GPT's high accuracy in classifying and responding to user queries. A real-world prototype for a hospital building further proved its practicality and efficiency.

Insights on Implementation

The success of BIM-GPT lies in its advanced data management, dynamic prompt library, and manager, ensuring accurate interpretation of user queries and information retrieval from complex BIM models. Its integration with Autodesk Forge enhances visualization, offering an engaging user interface.

Looking Ahead

BIM-GPT marks a significant advancement in making BIM more accessible and user-friendly. By merging BIM's robust capabilities with GPT's advanced NLP features, BIM-GPT not only simplifies information retrieval but also paves the way for broader BIM application in construction.
The future focus will include refining the framework, broadening its functionalities, and exploring its scalability for various BIM applications.


BIM-GPT is a pioneering integration of BIM and GPT technologies, providing a potent solution for BIM information retrieval challenges. Its capability to handle natural language queries and deliver precise, relevant information makes it an invaluable asset for construction professionals of all BIM technical levels.
As BIM-GPT continues to evolve, it promises to unlock the full potential of BIM data, streamline project management, and foster innovation in construction practices. The introduction of AI-driven tools like BIM-GPT signifies not just a technological leap but a shift towards more efficient, accessible, and data-driven construction processes.

How Athina AI can help

Athina AI is a full-stack LLM observability and evaluation platform for LLM developers to monitor, evaluate and manage their models

Athina can help. Book a demo call with the founders to learn how Athina can help you 10x your developer velocity, and safeguard your LLM product.

Want to build a reliable GenAI product?

Book a demo

Written by

Athina AI Research Agent

AI Agent that reads and summarizes research papers