EUIPO Trademarks Scraper — 100M+ EU & Global Trademarks (TMview) + AI Brand Clearance

⭐ Bookmark this actor → apify.com/dltik/euipo-trademarks-scraper — Apify ranks by bookmarks, so it directly drives the visibility of this scraper. One click. No signup beyond your existing Apify account.

Scrape 100M+ trademarks from EUIPO and 70+ national IP offices (USPTO, INPI France, DPMA Germany, WIPO Madrid, JPO Japan, UKIPO, CNIPA China and more) via the official TMview backend (tmdn.org) operated by the EU Intellectual Property Office. Filter by Nice class, status, office, application date, mark type — then go further than any other scraper: AI likelihood-of-confusion brand clearance, applicant → company resolution, free lifecycle & renewal signals, and a scheduled trademark-watch mode. No API key, no OAuth, no monthly quota — $0.01 per trademark, the disruptive sweet spot vs $0.10 elsewhere.

✨ What makes this EUIPO trademarks scraper different

Feature	This scraper	Other EUIPO scrapers
Price per result	$0.01	$0.05 - $0.10
Coverage (offices)	70+ IP offices via TMview	EUIPO-only
Full goods & services text (per Nice class)	✅ (enrichDetails)	❌ / extra fee / needs your OAuth
No OAuth / no API key (self-serve)	✅	often needs your EUIPO credentials
AI likelihood-of-confusion (brand clearance)	✅ (analyzeClearance)	❌
Free computed conflict score (phonetic + name + Nice)	✅ (referenceMark)	❌
Applicant → company record (SIREN + size + status)	✅ (enrichApplicant)	❌
Free lifecycle signals (days-to-expiry, opposition window, renewal-due)	✅	rare / paid
Trademark watch (only new/updated marks on a schedule)	✅ (incrementalMode)	❌
Nice + Vienna classification arrays	✅	partial
Concurrent detail enrichment (fast on big runs)	✅	❌
Success rate (live 30d)	100%	varies

What can this EUIPO trademark scraper do?

• 🇪🇺 100M+ live trademarks indexed by the EU IP Office's TMview (every active EUIPO trademark + 70+ national IP offices including USPTO, INPI, DPMA, WIPO Madrid)
• 🔎 Multi-criteria search — name, Nice class, status, office, mark type, application date range
• 🧠 AI brand-clearance / likelihood-of-confusion (analyzeClearance) — score every existing mark against a proposed brand name the way an EU examiner would: semantic (meaning) similarity, goods & services conflict, an overall confusion score, a verdict and a recommended action
• 🎯 Free computed conflict score (referenceMark) — even without AI, every result is scored against your proposed mark: Jaro-Winkler name distance + phonetic match (Metaphone/Soundex) + Nice-class overlap → a 0-100 computed_conflict_score
• 🏢 Applicant → company intelligence (enrichApplicant) — resolve the brand owner to a real company: French applicants to their SIREN + size + sector + active/ceased status, other countries to the matching national company-registry actor
• 📅 Free lifecycle & renewal signals — every mark returns lifecycle_stage, days_until_expiry, renewal_due_soon, and an opposition_open window with days left — no enrichment needed
• 👀 Trademark watch (incrementalMode) — schedule it and receive only NEW and UPDATED marks (competitor filings, status changes) since the last run
• 📝 Full goods & services text (enrichDetails) — the complete descriptive wording per Nice class, the field most EUIPO scrapers omit or hide behind your own OAuth
• 🎨 Nice + Vienna classification — numeric Nice classes plus Vienna image codes for figurative/combined marks
• 🚀 HTTP-only, no API key — TMview is a free public service from the EUIPO (Apify datacenter proxy is enabled by default to avoid edge IP filtering)

Brand clearance & likelihood of confusion — the premium wedge

Commercial trademark platforms (Corsearch, CompuMark, Markify) sell likelihood-of-confusion screening for $300–500 per mark per year. This actor puts the same core capability behind a single input flag.

Set referenceMark to the brand name you want to clear (and optionally referenceNiceClasses for the goods/services you'd use it in). Every result then carries a clearance object:

• Free tier (just referenceMark): mark_similarity (Jaro-Winkler), phonetic_match (Metaphone/Soundex), nice_overlap, computed_conflict_score (0–100) and a computed_conflict_band (high/medium/low/none).
• AI tier (analyzeClearance: true, paid plans): adds semantic_similarity (catches translations & synonyms — "SUN" vs "SOLEIL"), goods_conflict, an overall likelihood_of_confusion, a verdict (clear / caution / high_risk), a recommended_action and one-line reasoning.

⚖️ Screening triage to prioritise professional review — not legal advice.

What data can you extract per trademark?

Fields marked ⭐ require enrichDetails: true (billed as trademark-detail). 🏢 requires enrichApplicant. 🧠 requires referenceMark / analyzeClearance. Everything else comes from the base search at $0.01/result.

Field	Description
st13	TMview unique global ID (office code + sequential ID)
application_number	Office-level application number
mark_name	Verbal element of the mark
mark_type	Word, Figurative, Combined, 3-D, Sound, Colour, Hologram, Motion
office_code	Two-letter IP office (EM = EUIPO, FR = INPI, DE = DPMA, US = USPTO, WO = WIPO, ...)
status	Registered, Filed, Application published, Application opposed, Ended, Expired ...
lifecycle_stage	Normalized: registered / pending / expired / ended / unknown
is_active	true when the mark is registered or still in prosecution
days_until_expiry	Days from today to expiration (negative if lapsed) — populated with enrichDetails
renewal_due_soon	true when expiry is within 180 days
opposition_open / opposition_days_left	Whether the opposition window is still open + days left
application_date	YYYY-MM-DD
nice_classes	Array of Nice class numbers ([9, 38, 41, 42])
vienna_codes	Array of Vienna image classification codes (figurative/combined marks)
applicant_name / applicant_names	Primary applicant + full list
territories_protected	Country codes where the mark is enforced (Madrid + EU)
mark_image_url / mark_thumbnail_url	TMview-hosted mark image + thumbnail
tmview_url / office_record_url	Deep links to TMview + the originating IP office
scraped_at	ISO-8601 UTC timestamp
⭐ registration_number / registration_date / expiration_date	Office registration data
⭐ opposition_period_start / opposition_period_end / opposition_deadline	Opposition window dates
⭐ status_detail / status_date / kind_mark	Detailed status + Individual/Collective/Certification
⭐ goods_and_services	Array of {nice_class, description} — full descriptive wording per class
⭐ applicants_detail	Array of {name, country, address}
⭐ representatives / renewals	Legal representative(s) + renewal history
🏢 applicant_company	Resolved French company: siren, legal_name, naf_code, sector, employee_range, is_active, risk_flag
🏢 registry_hint	For non-FR applicants: pointer to the matching national company-registry actor
🧠 clearance	{mark_similarity, phonetic_match, nice_overlap, computed_conflict_score, computed_conflict_band} + AI {semantic_similarity, goods_conflict, likelihood_of_confusion, verdict, recommended_action, reasoning}

How to scrape EUIPO trademarks in 5 steps

1. Create a free Apify account — no credit card required
2. Open EUIPO Trademarks Scraper on Apify Store
3. Enter query — brand name or keyword (apple, louis vuitton, tesla)
4. Optionally filter — offices=['EM'], niceClasses=[9, 42], statuses=['Registered']; set enrichDetails=true for full dates + goods & services, enrichApplicant=true for company data, referenceMark + analyzeClearance=true for AI clearance
5. Click Start — clean JSON in seconds, exportable to CSV / JSON / Excel

How much does it cost to scrape EUIPO trademarks?

PAY_PER_EVENT — $0.01 per trademark result ($10 per 1,000). Failed/empty runs are not charged. Optional add-ons are OFF by default and billed only when you enable them:

Event	When	Price
trademark-result	every mark (identity, status, Nice/Vienna, applicant, + free lifecycle & conflict score)	$0.01
trademark-detail	enrichDetails — reg/expiry/opposition dates + full goods & services + representatives + renewals	+$0.01
applicant-enriched	enrichApplicant — only for applicants resolved to a real company (FR SIREN)	+$0.005
clearance-analyzed	analyzeClearance — AI likelihood-of-confusion per mark (paid plans)	+$0.02

Still far under the $0.05–$0.10 competitors charge for less data — and none of them offer AI clearance, applicant intelligence, or watch mode at any price.

Run size	Trademarks	Base ($0.01)	+ Detail ($0.02)	+ AI clearance ($0.04)
Quick test	30	$0.30	$0.60	$1.20
Standard	300	$3.00	$6.00	$12.00
Deep	1,000	$10.00	$20.00	$40.00

Input parameters

Parameter	Type	Default	Description
query	string	—	Brand name / keyword search (required)
offices	string[]	[] (all)	Office codes: EM, FR, DE, US, WO, JP, UK, ...
statuses	string[]	[] (all)	Registered, Filed, Expired, Application published, ...
niceClasses	int[]	[] (all)	Nice class numbers 1-45
tmTypes	string[]	[] (all)	Word, Figurative, Combined, 3-D, Sound, Colour
applicationDateFrom / applicationDateTo	string	—	YYYY-MM-DD window
enrichDetails	boolean	false	Full detail per result (dates, goods & services, representatives, renewals). +$0.01/result. Fetched concurrently.
enrichApplicant	boolean	false	Resolve applicant → company (FR SIREN live; other countries → registry pointer). +$0.005/resolved.
referenceMark	string	—	Proposed brand name to clear → free computed conflict score on every result
referenceNiceClasses	int[]	[]	Intended Nice classes — sharpens the clearance score
analyzeClearance	boolean	false	AI likelihood-of-confusion vs referenceMark (paid plans). +$0.02/mark.
referenceGoods	string	—	Free-text intended goods/services, for AI clearance context
incrementalMode	boolean	false	Watch mode: only NEW/UPDATED marks on recurring runs
includeUnchanged / includeExpired	boolean	false	Also emit unchanged marks / expired markers (watch mode)
stateKey	string	default	Namespace for the watch baseline
detailConcurrency	integer	4	Parallel detail requests (1-8) when enrichDetails is on
pageSize	integer	30	TMview page size (1-200)
maxResults	integer	30	1-5000
proxyConfig	object	datacenter	Apify datacenter proxy enabled by default

Output example

Record with enrichDetails + enrichApplicant + referenceMark: "APPEL" + analyzeClearance:

{
  "st13": "EM500000000000753",
  "mark_name": "APPLE",
  "mark_type": "Word",
  "office_code": "EM",
  "status": "Registered",
  "lifecycle_stage": "registered",
  "is_active": true,
  "days_until_expiry": 3562,
  "renewal_due_soon": false,
  "opposition_open": false,
  "application_date": "1996-04-01",
  "registration_date": "1999-02-03",
  "expiration_date": "2036-04-01",
  "nice_classes": [9, 16, 38, 41, 42],
  "goods_and_services": [
    {"nice_class": 9, "description": "Computers, computer terminals, keyboards ..."}
  ],
  "applicant_name": "Apple Inc.",
  "applicant_company": null,
  "registry_hint": {"country": "US", "registry_actor": "usa-ny-company-registry-scraper", "registry_actor_url": "https://apify.com/dltik/usa-ny-company-registry-scraper"},
  "clearance": {
    "reference_mark": "APPEL",
    "mark_similarity": 0.907,
    "phonetic_match": true,
    "nice_overlap": [9, 42],
    "computed_conflict_score": 88,
    "computed_conflict_band": "high",
    "semantic_similarity": 20,
    "goods_conflict": 90,
    "likelihood_of_confusion": 82,
    "verdict": "high_risk",
    "recommended_action": "avoid",
    "reasoning": "Near-identical spelling and phonetics with heavy Nice-class 9/42 overlap makes confusion likely."
  },
  "tmview_url": "https://www.tmdn.org/tmview/#/tmview/detail/EM500000000000753",
  "scraped_at": "2026-07-01T12:32:01Z"
}

Use cases — trademark watch, brand clearance & IP intelligence

• 🧠 Brand clearance / naming — before you file, run your candidate name as referenceMark with analyzeClearance and get every conflicting mark ranked by likelihood of confusion across EU + national offices
• 👀 Trademark watch services — schedule incrementalMode on a watchlist and receive only newly-filed or status-changed marks (competitor look-alikes) each day/week
• 🏢 Brand-owner lead-gen & competitive intel — enrichApplicant turns "who owns this mark" into a full company profile (SIREN, size, sector, active/ceased) — impossible with a trademark-only scraper
• ⚖️ IP attorney prior-art search — pull every trademark named X across EU + national offices in seconds, ranked by Nice class, with full goods & services text
• 🚨 Opposition & renewal alerts — surface every mark whose opposition_open window or renewal_due_soon flag fires within your horizon
• 📊 IP portfolio analytics — aggregate application_date × office × Nice class × applicant to map filing trends and competitor portfolios

Use the EUIPO scraper via API

import requests

run = requests.post(
    "https://api.apify.com/v2/acts/dltik~euipo-trademarks-scraper/run-sync-get-dataset-items",
    headers={"Authorization": "Bearer YOUR_APIFY_TOKEN", "Content-Type": "application/json"},
    json={
        "query": "lumina",
        "offices": ["EM", "FR"],
        "referenceMark": "lumina",
        "referenceNiceClasses": [9, 42],
        "analyzeClearance": True,
        "enrichDetails": True,
        "maxResults": 50
    }
).json()
print(f"Scraped {len(run)} trademarks")

FAQ

Is this an official EUIPO API? TMview (tmdn.org) is the official EU Intellectual Property Office trademark search backend, aggregating EUIPO + 70 national IP offices. This scraper hits the same public JSON endpoint the TMview web UI calls — no HTML parsing, no anti-bot circumvention.

How is the AI likelihood-of-confusion different from the free conflict score? The free computed_conflict_score (from referenceMark) is string-level: Jaro-Winkler distance + phonetic keys + Nice-class overlap. The AI layer (analyzeClearance) adds meaning: it catches translations, synonyms and conceptual overlap ("SUN"/"SOLEIL"), and it reasons about goods & services conflict the way an EU examiner does. Use the free score to triage cheaply, then AI on the shortlist. It's screening triage, not legal advice.

What's the difference vs the nexgendata/euipo-esearch-trademarks actor? Price ($0.01 vs $0.10), coverage (70+ offices vs EUIPO-only), and depth: we return goods & services, opposition data, applicant-company resolution, lifecycle signals and AI clearance — none of which it offers. And we're self-serve (no BYO EUIPO OAuth, unlike the actors that do return goods & services).

Will TMview rate-limit my runs? TMview has a soft ~3-5 req/s limit. Detail enrichment runs through a small concurrent pool with exponential backoff on 429, so even a 5,000-mark enriched run finishes comfortably inside the actor timeout. Datacenter proxy is enabled by default.

Does the watch mode charge me for everything every run? No. In incrementalMode only NEW and UPDATED marks are emitted and billed; unchanged marks are skipped (unless you set includeUnchanged). That's what makes daily brand monitoring cheap.

---

⭐ Found this EUIPO Trademarks Scraper useful? Bookmark it — Apify ranks actors by bookmarks, so it's the strongest single signal for Store visibility. One click.

The dltik IP intelligence suite

Trademarks are one right — cover the whole IP + company graph with the rest of the suite:

Actor	What it does	Price
Espacenet Patents Scraper	130M patents from EPO (Espacenet)	$0.015/patent
USPTO Patents Scraper	US patents via PatentsView	$0.01/patent
Pappers Sirene Scraper	26M French companies (SIREN/SIRET) — powers enrichApplicant	$0.008/result
INPI MCP Server	French IP (patents/trademarks/designs) for AI agents	$0.01/tool-call
BODACC Scraper	French commercial court announcements	$0.003/record
SEC EDGAR MCP Server	US public companies for AI agents	$0.01/tool-call
TED Europa Scraper	EU public procurement tenders	$0.005/tender
Pappers MCP Server	French company data for Claude / GPT	$0.01/tool-call

License: MIT · Author: dltik