Scento Perfume Compendium
Free perfume database API for AI assistants over MCP and REST.Structured perfume data with no key and no signup.
Highlights
What you can do with it
- Ask Claude for woody office fragrances and get a real list with notes and families.
- Build a perfume search mini program filtered by brand, scent and ingredient.
- Check where sandalwood or iris appear across different perfumes for review writing.
- Teach a class how a language model calls an external knowledge tool.
How to use it
- 1 Open the Scento developer page to see the MCP and REST options.
- 2 For Claude or another MCP client, add the Scento server address in the MCP configuration; no key is needed.
- 3 For REST, send HTTP requests to query perfumes, ingredients or families.
- 4 Define the fields your application needs, such as name, notes, family and ingredients.
- 5 Check the returned data before using it for recommendations or display.
- 6 Extend from there into recommendation engines, shopping guides or Q&A bots.
Who it is for
Pros
- No registration and no API key, so setup takes minutes.
- Structured output designed for model tool calls rather than raw HTML.
- Both MCP and REST access fit different tech stacks.
Limitations
- The catalogue is limited to 1,400 perfumes and 400 ingredients.
- Free with no key also means no published quota or uptime commitment.
- Family and note data describe composition, not wear performance.
What Scento Perfume Compendium is
Scento Perfume Compendium is a perfume database connector from Scento, offered through an MCP server and a REST API for use by AI assistants such as Claude and ChatGPT. It exposes data on 1,400 designer and niche perfumes, 400 fragrance ingredients, and a fragrance family chart, with no registration and no API key required. The idea is to give a language model real structured facts about specific perfumes, accords and ingredients so that answers come from the catalogue rather than from memory, which reduces invented detail. The response format is shaped for tool calls, so it drops into a conversation or an application without much parsing, and the family chart lets recommendations move naturally from one family to its neighbours, for example from oriental to amber, woody or spicy. The underlying data comes from Scento's own retail catalogue and industry knowledge, and the service is currently free.
What you can do with it
- Query perfumes by brand, name, notes, family or occasion.
- Look up what an ingredient does and where it appears.
- Extend recommendations along family relationships.
- Build recommendation engines, shopping guides or Q&A bots on the REST API.
Who it is for
- Developers adding domain knowledge to a chatbot.
- Perfume e-commerce and content teams building guides.
- Perfumers and reviewers researching formulas and notes.
What to watch out for
The catalogue is smaller than a full fragrance database at 1,400 perfumes, so newer or very small releases will be missing, and ingredient data stops at 400 entries. It is free with no key and no published quota, which means there is no service guarantee; if you depend on it in production, cache responses and plan for downtime or change. Notes and families describe composition, not how a fragrance wears on you, so recommendations still need human judgement and testing on skin. Check Scento's terms before reselling or redistributing the data inside your own product.
FAQ
Do I need an API key?
No. The service requires no registration and no API key for either MCP or REST access.
What data is included?
1,400 designer and niche perfumes, 400 fragrance ingredients and a fragrance family chart.
Which clients are supported?
MCP clients such as Claude and ChatGPT, plus any platform that can call the REST API.
Is it free?
It is currently free, but with no published quota, so avoid building a critical dependency without caching and a fallback.
Why it is free
What the official page says: 免费
Free limits, trial length and features are decided by the official site and can change at any time.