BioLM MCP Server

The BioLM MCP (Model Context Protocol) server lets you connect AI assistants like Claude and ChatGPT directly to BioLM's computational biology tools — no API code required. Once connected, you can run protein folding, antibody design, variant effect prediction, and more by simply describing what you want in conversation.

What You Can Do

The BioLM MCP server exposes 20+ tools across core computational biology workflows:

Category Tools
Protein structure ESMFold, ABodyBuilder3, NanobodyBuilder2
Sequence embeddings ESM-2 (150M), ESM-C (300M, 600M)
Variant effect ESM-1v (ensemble), Evo2 1B Base
Antibody annotation & design SADIE, AbLang-2, IgBert, AntiFold
Nanobody tools nanoBERT, NanobodyBuilder2
Synthetic biology DNA Chisel, Evo2
Safety & stability BioLMTox-2, TEMPRO 650M

Connecting to the MCP Server

The server is available at https://biolm.ai/mcp. There are two ways to authenticate:


OAuth 2.0 — recommended for personal use. Add the server URL to your AI client and it will redirect you to BioLM's login page automatically. No API key management needed.

API token — recommended for teams and programmatic access. Pass your BioLM API key as a header:

X-BioLMAI-Token: YOUR_API_KEY

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json :

{
    "mcpServers": {
        "biolm": {
            "type": "http",
            "url": "https://biolm.ai/mcp",
            "headers": {
                "X-BioLMAI-Token": "YOUR_API_KEY"
            }
        }
    }
}

Restart Claude Desktop — you should see biolm appear in the tools panel.

Claude Code

Add to ~/.claude.json , or via CLI:

claude mcp add biolm --transport http https://biolm.ai/mcp \
--header "X-BioLMAI-Token: YOUR_API_KEY"

Claude.ai (Web)

Go to Settings → Integrations and add https://biolm.ai/mcp . Claude.ai will initiate the OAuth flow automatically.

ChatGPT

In Settings → Tools → Add Tool, enter https://biolm.ai/mcp  and set the auth header X-BioLMAI-Token .

Local install (npm)

This option opens a browser login and saves credentials for future sessions.

npx @biolm/biolm-mcp --login

Trying It Out

Once connected, ask your AI assistant something like:

"Fold this protein sequence and tell me the mean pLDDT score."

The assistant will call the appropriate BioLM tool, return results, and you can iterate from there — all in conversation.

Filtering Tools

If you only need a subset of tools, the server URL supports query parameters:

https://biolm.ai/mcp?tag=antibody https://biolm.ai/mcp?tool=predict_v3_esmfold
Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.

Still need help? Contact Us Contact Us