Looking for a free perfume API that works with AI assistants? Scento Perfume API is a fragrance API and perfume MCP server over Scento's verified perfume database: 1,500+ perfumes, 160+ houses, 400+ raw materials, every notes pyramid traced to the house or an authorized retailer.
What you get: a perfume notes database you can query by name, note, family or occasion; a fragrance ingredients database (400+ raw materials with the perfumes that use them); similar-scent and head-to-head comparison endpoints; scent family maps; occasion edits and wardrobe building; a perfumery lexicon; and a journal search for editorial context. Twelve tools in total: perfume-search, perfume-portrait, scent-neighbours, perfume-contrast, note-portrait, material-match, house-portrait, family-map, occasion-edit, wardrobe-composer, perfumery-lexicon, journal-search.
Use it two ways. As an MCP server (remote, streamable HTTP) at https://www.scento.com/mcp, or via npx -y scento-perfume-api and uvx scento-perfume-api for stdio clients. Or as plain REST at https://www.scento.com/api/perfume/v1 with an OpenAPI document at /openapi.json. Both are free and need no API key.
Typical uses: a chatbot that recommends perfumes with real notes, a fragrance finder, a "smells like" lookup, a retailer enriching product pages with a notes pyramid, a student project, or an AI agent that needs a trustworthy perfume database API instead of guessing. Responses come in 10 languages.
Maintained by Scento. Docs and live examples: https://www.scento.com/developers
Connect
Add one remote server, no key needed.
Claude (web or desktop): Settings → Connectors → Add custom connector → paste https://www.scento.com/mcp.
ChatGPT: Settings → Apps → Advanced → developer mode → Create → paste https://www.scento.com/mcp, authentication None.
Cursor, VS Code, Windsurf and other clients: add {"scento-perfume-api": {"type": "http", "url": "https://www.scento.com/mcp"}} to mcpServers.
Clients that only run local servers: npx -y scento-perfume-api (Node) or uvx scento-perfume-api (Python).
Prefer plain HTTP? The same answers live at https://www.scento.com/api/perfume/v1, described in https://www.scento.com/api/perfume/v1/openapi.json.
Links
コメント