By Aaron Tay, Head, Data Services
SMU IITS recently launched SMU Studio Chat, a trusted platform for secure AI adoption.
Currently in pilot testing, SMU Studio Chat is available to staff, faculty, and researchers. It provides access to a variety of models from DeepSeek, Google, OpenAI and Anthropic.
This librarian’s personal view after using it
SMU Studio Chat is based on LibreChat, an open-source AI platform that has been used by universities such as Harvard and Stanford. Like many modern LLM chat platforms, it supports features such as web search, file uploads, artifact creation, memory, and connections to MCP servers.
It supports a range of models, including the image-generation model Flux.2 Pro.
More advanced users may also appreciate the ability to create their own skills, agents and sub-agents.
I have not yet tested this feature enough to compare it with a fully agentic coding environment such as Claude Code CLI or Claude Code Desktop, but it certainly looks interesting.
Another important point is that SMU Studio Chat runs on SMU’s own Azure environment. This means that files you upload remain on SMU servers rather than being sent to public platforms such as OpenAI or Anthropic. This is particularly important because many publisher terms prohibit uploading licensed content to public AI services.
Overall, SMU Studio Chat looks like a strong option. My main complaint is that it is fairly easy to run into monthly quotas if you use the heaviest models all the time and you can’t add any MCP servers beyond what is there by default .
Academic search in SMU Studio Chat
You may have read about connecting ChatGPT or Claude to academic MCP servers such as Scite and Consensus, allowing the chatbot to search scholarly sources through those services.
Can you do the same in SMU Studio Chat?
Thanks to our excellent colleagues at SMU IITS, the answer is yes. SMU Studio Chat has been set up so that users can connect to both Scite and Consensus through MCP.
First, go to the MCP server settings in the sidebar. From there, you can connect to Consensus and/or Scite.
Concluding thoughts
SMU Studio Chat is not just another chatbot interface. For SMU staff, faculty and researchers, its value lies in combining access to strong AI models with a more secure institutional environment and the ability to connect to research-focused tools such as Scite and Consensus through MCP.
That said, users should still treat the results carefully. MCP-enabled academic search can help surface papers and provide useful starting points, but it does not replace careful searching, reading, appraisal or citation checking. As with all AI-assisted research tools, the quality of the answer depends on the quality of the sources retrieved and how well the model uses them.
For researchers who want to experiment with AI-supported literature discovery without uploading content to public AI platforms, SMU Studio Chat is well worth trying. The addition of academic MCP servers makes it even more useful, especially for those who want a convenient way to combine conversational AI with scholarly search tools.
You will usually be asked to sign in to your Consensus or Scite account, if you have not already done so, and then confirm the connection.
Once the MCP servers are set up, you can ask SMU Studio Chat to search for academic papers. The LLM will usually decide whether to use Consensus, Scite, or both as tools to retrieve scholarly sources.
