ChatGPT guide

Search PubMed papers with ChatGPT

Count and list biomedical papers on a topic, with titles, journals and dates. PubMed is free.

1. Connect looot to ChatGPT

Open ChatGPT Plugins at chatgpt.com/plugins.
Select the plus button, then "Add custom MCP server".
Enter a name (looot) and a short description, for example: live data APIs, paid per call.
Under Connection, choose Public endpoint and paste https://api.looot.ai/mcp, including the /mcp path.
Configure authentication, read the risk warning and select "I understand and want to continue".
Select "Create as a plugin", review the tools ChatGPT found on the server, then install the plugin.
looot's sign-in page opens when ChatGPT first needs your account: sign in and allow access. Runs are paid from your prepaid looot balance.

OpenAI's guide does not say which ChatGPT plans can add a custom MCP server. It says account and workspace policies apply to adding and using custom MCP servers.

Source, read 2026-11-10: OpenAI developer docs: Connect and test your plugin

2. Ask ChatGPT

Start a new conversation, type @ in the prompt box and select the looot plugin. Then paste the prompt.

Prompt

Use looot to search PubMed for [topic] with the job research.papers.pubmed.search, asking for 10 results. Then run research.papers.pubmed.summary on the ids it returns. Tell me how many papers match, then list each with title, journal and publication date.

3. Jobs it calls and what they cost

research.papers.pubmed.search
1 call, from free (PubMed (NCBI E-utilities), the cheapest of 1 provider)
About the cheapest provider
research.papers.pubmed.summary
1 call, from free (PubMed (NCBI E-utilities), the cheapest of 1 provider)
About the cheapest provider

One use costs from free at today's cheapest providers. A provider that misses can fall back to the next one, which can cost more; looot charges only the calls that ran, from your prepaid balance.

Providers in the real run below:PubMed (NCBI E-utilities)

4. A real run

We ran this task through the looot MCP on 2026-11-10. Asked: semaglutide[ti] AND 2025[dp]. Total charged: $0.

906 papers have semaglutide in the title with a 2025 publication date. The search returned the first 10 PubMed ids.
The summary call turned ids into titles, for example: "Efficacy of Subcutaneous Semaglutide Dose Escalation in Reducing Insulin in Patients With Type 2 Diabetes" (Federal Practitioner, November 2025).
The other two we read: a case series on liver biomarkers in adolescents (JPGN Reports) and a mouse study of nephritis (Kidney360).

research.papers.pubmed.search on pubmed-search (PubMed (NCBI E-utilities)): answered, charged $0. Run id run_94ccc006b0d7482b86b5a8d809e689a1.
research.papers.pubmed.summary on pubmed-summary (PubMed (NCBI E-utilities)): answered, charged $0. Run id run_3541f410c80d4db0a1ca3d1d1ea07301.

Good to know

PubMed search terms accept field tags: semaglutide[ti] searches titles only, 2025[dp] limits to a publication year.

The same task in Claude All ChatGPT guides

Give your agent every data provider, one key.

Start your workspace, top up, and paste one prompt into the agent you already use.

Start your workspace See what agents can do