01
Ask like a human
Beast-mode winter oud. Office-safe summer. Something like Aventus but quieter. Your agent understands the intent.
natural language → tool selection
Fragrance MCPPerfume MCP126k+ scents
Perfume MCP with a sense of smell
Free fragrance MCP and perfume MCP server for AI agents. Query 126k+ scents for longevity, sillage, seasonality, and community scores from Cursor or Claude.
›
Full profile for Creed Aventus
Aventus
Creed · 8,420 wear votes
★ 4.5sample
longevity
long lasting · high confidence
0.88
sillage
strong · high confidence
0.84
appreciation
beloved · high confidence
0.91
fruitysweetwoodymuskysmoky
get_fragrance_profile200 OK · 84ms
What this is
A perfume MCP, a fragrance MCP, and a cologne catalog
scentrev-mcp is a free perfume MCP server and fragrance API. One endpoint covers 126,000+ scents and 7,500+ brands. A cologne MCP is the same catalog with US wording for men's fragrance. Agents read scores for longevity, sillage, and notes instead of guessing from a webpage.
- Perfume MCPThe same server, written the way US searchers ask for a perfume MCP server.
- Fragrance MCPThe industry name for this catalog: tools for notes, wear, and comparisons.
- Perfume dupe finderCheaper similars via find_dupes, then a side-by-side compare.
- Perfume recommendation APIFilter by season, longevity, and sillage instead of adjectives.
Depth you can actually query
Every filter and recommendation draws on structured notes, accords, perfumers, and community similarity. Real signals, not vibes.
0k+
Fragrances
0+
Brands
0+
Perfumers
0
Notes
0+
Accords
0
MCP tools
Built in India
India's first MCP for perfume data
ScentRev MCP is the first Model Context Protocol server in India dedicated to structured fragrance intelligence: longevity, sillage, seasonality, accords, and community scores your agents can query in one call. Read the perfume MCP guide to install it in Cursor or Claude.
- 126k+ fragrances, normalized for LLMs and REST
- Drop into Cursor, Claude Desktop, or your own stack
- Free during the current period. No credit card
Drops into any MCP client
One call, the whole story of a scent
Community knowledge, normalized into metrics your agents and apps can trust.
Full fragrance profiles
Longevity, sillage, seasonality, gender lean, sentiment, accords, and similar fragrances. One structured response, no scraping.
longevity
0.82
sillage
0.71
appreciation
0.79
MCP-native
Drop into Cursor or Claude Desktop with one config block.
Normalized scores
Every metric is 0 to 1 with an n_records confidence count. No raw vote dumps.
REST or MCP, same scoring model
Call it from an agent or hit the endpoints directly. You get the same normalized data either way.
- API key or OAuth. Same tools either way.
- Predictable JSON, typed fields
- Confidence counts on every metric
{
"mcpServers": {
"scentrev-mcp": {
"url": "https://api.scentrev.com/mcp/",
"headers": {
"Authorization": "Bearer frag_live_..."
}
}
}
}Ask in plain language, get structured data back
Best beast-mode winter scent from Lattafa?Full profile for Creed AventusOffice-safe summer fragrances under $80What smells like Tom Ford Oud Wood?Eternal longevity oud for menTop accords in Jazz ClubIs Club de Nuit Intense beloved?Notes in Lost Cherry
22 tools, one fragrance brain
Search, filter, and pull structured profiles from community data.
01
search_fragrances
Free-text search across the catalog
02
search_fragrances_filtered
Filter by brand, season, notes, discovery sort
03
search_by_references
Blend note pyramids from 2+ reference scents
04
list_brands
Browse brands with fragrance counts
05
resolve_slug
Resolve informal names to slugs
06
compare_fragrances
Side-by-side wear and value metrics
07
get_fragrance_profile
Every metric for one fragrance
08
get_wear_summary
When and where it shines
09
get_identity
Gender lean and character
10
get_performance
Longevity and sillage scores
11
get_appreciation
Community sentiment and love score
12
get_reminds_of
Closest-smelling alternatives
13
get_note_pyramid
Top, heart, and base notes
14
get_accords
Dominant accords, weighted
15
get_scent_card
One call for wear, sentiment, and similars
16
suggest_occasion
Shortlist for office, date, summer, winter
17
find_dupes
Cheaper adjacent alternatives
18
blind_buy_risk
How risky a blind buy is
19
decode_taste
Shared DNA across loved scents
20
wardrobe_add
Save a bottle to your wardrobe
21
wardrobe_list
List saved wardrobe rows
22
wardrobe_analyse
Spot gaps in a saved wardrobe
From question to structured scent intel
Three steps your agent already knows how to take.
02
Score like a critic
Longevity, sillage, seasonality, and community love as 0 to 1 metrics with vote counts. No scraping, no guesswork.
get_fragrance_profile · normalized scores
03
Ship like a builder
Same data over MCP or REST. One bearer key. Structured JSON your product can trust.
Cursor · Claude · HTTP
Connected in under a minute
1
Create a key
Sign up and generate a key in the dashboard. No credit card.
2
Add the config
Paste one block into Cursor or Claude Desktop. That is the whole setup.
3
Ask your agent
Your agent now answers fragrance questions with structured, sourced data.
Questions, answered
Everything you need to know before wiring scentrev-mcp into your stack. Still stuck? The docs go deeper.
Start building in minutes
Sign up, generate a key, connect your MCP client. Free tier with daily and weekly usage limits — no credit card.