Model Context Protocol · Live
Supplement intelligence,
as a tool for your agents.
2,999 scored ingredients. 620 analyzed products. 2,400+ interactions. Ship it to Claude, Cursor, or any MCP-compatible client in 30 seconds.
The 6 tools your agent gets
search_ingredients
Search the IngredientMD database of 2,999 supplement ingredients by name or category. Returns ranked matches with their evidence-based score, hook (one-line summary), and URL. Use this first to find ingredients before calling get_ingredient for full detail.
get_ingredient
Get full detail for one ingredient by slug. Returns name, score, verdict, clinical dose, evidence grade, primary benefits, mechanisms, caution flags, and rich content (FAQs, origins, dose-response, food sources). Use after search_ingredients to pull authoritative data on a specific ingredient.
check_interactions
Detect conflicts (absorption problems, dangerous combinations) and synergies (pairs that amplify each other) across a supplement stack. Uses IngredientMD's 2,400+ interaction database. Returns severity-ranked issues with actionable timing guidance (e.g., 'take 2h apart').
score_stack
Run IngredientMD's AI scoring engine on a supplement stack. Returns an overall 1-10 score, verdict (EXCEPTIONAL to AVOID), stack purpose, strengths, weaknesses, timing/food guidance, and who benefits vs should avoid. This is the same analysis shown on /stack for end users. Costs 1-2 credits per call.
find_top_products
Find the top-scored products in IngredientMD for a given category. Returns products with their AI-verified score, brand, and purchase URL. Use this when a user asks "what's the best X?" — we've already scored 620 popular products.
get_dose_response
Get the dose-response curve for an ingredient: minimum effective dose, optimal dose, upper safe limit, and confidence level. Source: IngredientMD's scientifically-sourced clinical database covering 236+ ingredients.
Free · No credit card
Mint your trial key.
100 calls/month. Read-only MCP tools — search, get, interactions, dose-response. Shown once, stored hashed.
Tiers
Trial
Free
100 calls/month
Read-only tools. Perfect for prototyping.
5 tools
Starter
$9/mo
1,000 calls/month
All tools including AI-backed score_stack.
6 tools
Growth
$29/mo
10,000 calls/month
Priority support + SLA commitment.
6 tools
Enterprise
Custom
1M calls/month
Dedicated key, custom tools, co-branded.
6 tools
Starter+ includes score_stack — our AI-backed scoring. Trial keys can't call it (it's compute-heavy on our end). Email partnerships@ingredientmd.com to upgrade.
Drop these into your client
Claude Desktop
~/Library/Application Support/Claude/claude_desktop_config.json{
"mcpServers": {
"ingredientmd": {
"url": "https://ingredientmd.com/api/mcp",
"headers": { "Authorization": "Bearer imd_..." }
}
}
}Cursor
.cursor/mcp.json{
"mcpServers": {
"ingredientmd": {
"url": "https://ingredientmd.com/api/mcp",
"headers": { "Authorization": "Bearer imd_..." }
}
}
}Cline / Continue
settings.json or mcp.json{
"mcp.servers": {
"ingredientmd": {
"url": "https://ingredientmd.com/api/mcp",
"auth": { "type": "bearer", "token": "imd_..." }
}
}
}