No: there is no public, documented TMview API, and EUIPO's API Portal doesn't list one. The official way to get EU trade mark data into code is EUIPO's own Trademark search API, which covers the EU register only; for the national registers TMview adds, there is no single official API. Multi-office search from code means one integration per office or a third-party trademark API, which the second half of this guide shows.
Everything below is as of November 13, 2026, with a source on each claim. If you want to search TMview by hand, the TMview search guide walks through the interface.
Does TMview Have an API?
TMview is EUIPO's free search tool for the TMDN network. It searches 143,842,303 trade marks from 81 participating offices (counts captured November 4, 2026). It is a genuinely useful public tool, and its footer is clear about what it is not: TMview and DesignView are not official registers, and the information in them has no legal effect.
The place an API would live is EUIPO's API Products catalogue. As of November 13, 2026 it lists 11 products: Trademark search 1.1.0, Design search 1.1.0, EUTM Filing 1.1.0, Design filing 1.1.1, Goods And Services 1.2.0, Me 1.2.0, My EUIPO mailbox 1.2.1, Document Repository 1.0.0, Persons 1.0.0, Product Indications 1.0.0 and Out of Commerce Works 1.2.0. None of them is TMview.
Two terms come up throughout. An EUTM (EU trade mark) is a single registration covering the whole European Union, filed with EUIPO. An international registration is filed through WIPO under the Madrid Protocol, which lets one application designate many countries; each designated country is protected separately.
Here are the realistic options:
| Option | Covers | Access | Good for |
|---|---|---|---|
| TMview web app | 81 offices | Free, manual; export capped at 150 marks, CAPTCHA on downloads | Occasional manual checks |
| EUIPO Trademark search API | EUIPO register (EUTMs, EU designations) | Free, OAuth, approval plus ID documents | Official EU data in code |
| National office APIs | One office each | Varies by office | One market |
| WIPO | Madrid international registrations | No search API; paid bulk XML | Madrid data at scale |
| Third-party trademark API, e.g. Signa | 15 offices | API key, credits | Multi-office search in code |
Can You Call TMview's Internal API Directly?
The TMview web app does fetch JSON, and the requests show up in any browser's network tab during a search. Results come from internal endpoints under https://www.tmdn.org/tmview/api/: a POST to /tmview/api/search/results for the hit list, /api/search/filter/* for the filter sidebar and /api/general/* for configuration. These are not offered as a product. They have no documentation and no public versioning, and they can change without notice. One path in the same app is literally named /tmview/api/internal/search/version.
The JSON is easy to read, which is why people ask. A results page carries fields such as ST13 (the WIPO ST.13 application identifier, for example SE500000200007385), tmName, tmOffice, tProtection (territories), applicationNumber, tradeMarkStatus (values like Registered, Filed, Ended), niceClass, applicantName, applicationDate and thumbnail URLs. A word search for OATLY returned 258 results over 9 pages of 30.
Calling that endpoint from a script, outside the browser app, was refused. A direct request from curl on November 13, 2026 got back an HTML "Request Rejected" page with a support ID, which is a web application firewall response. The configuration the app loads also has reCaptcha enabled (reCaptchaEnabled: true).
The terms say the same thing in legal language. TMview's "Legal notice" link goes to the EUIPN legal notices, operated by EUIPO. Access is "granted for consultation", and users may not do anything that harms EUIPO's network performance or security. Extracting substantial parts of EUIPO's databases, however it is done, is forbidden. EUIPO also opts out of any copyright exception for text and data mining or web scraping, citing Article 4(3) of Directive (EU) 2019/790, whether done "by any means, including bots, scrapers, or other automated processes", and expresses that machine-readably as <TDM-RESERVATION: 1>.
There is one narrow exception: text and data mining for scientific research, on conditions (results can't be passed on or used for other purposes without EUIPO's prior written consent, and are destroyed when no longer needed). Commercial users are encouraged to ask EUIPO for specific, explicit permission. The notice adds that partner offices' data is displayed as received, without guarantee.
Taken together, scraping TMview or building a product on its internal endpoints carries a legal risk and an availability risk. This is a summary of what the terms say, not legal advice; consult a trademark attorney for guidance on your situation.
What TMview Gives You Without an API
Before reaching for code, check whether TMview's own features cover the job. From the configuration the web app loads (November 13, 2026):
- Export, not bulk download: up to 150 selected trade marks, with downloads behind a CAPTCHA. Results come in pages of 30.
- Saved searches (sign-in required): up to 50, each holding up to 5,000 results. TMview notifies you when new marks enter the results or existing ones drop out.
- Alerts (sign-in required): up to 50 marks you pick, with notices on status changes, the end of an opposition period, expiry, owner changes and more.
Notifications arrive in the app and by email. If you wanted an API only to watch a handful of names, this may be all you need, and it costs nothing.
Official Alternatives to a TMview API: EUIPO, National Offices and WIPO
EUIPO. The Trademark search API is free, uses OAuth, and is approval-gated: you test in a sandbox, then apply for production with identity documents. It covers EUIPO's register only, meaning EUTMs and international registrations designating the EU, not the national registers TMview also shows. Its query language matches exact values and asterisk wildcards, and the spec lists no fuzzy or phonetic operator. The EUIPO API guide covers access, OAuth and queries step by step; the EUIPO office page summarises the register.
National offices. TMview's data comes from each participating office, which the legal notice calls partner data-holders. Some of those offices publish their own APIs or bulk data; the USPTO does, covered in the USPTO trademark API guide. Rebuilding TMview this way means one integration per office, each with its own authentication, schema and update cadence.
WIPO. WIPO offers no public search API for the Global Brand Database or Madrid Monitor, and the Global Brand Database terms forbid automated queries, scraping and bulk downloading; Madrid Monitor bulk data is sold separately as XML, and the WIPO trademark search guide has the detail.
Multi-Office Search Through Signa's API
Signa is a trademark API that normalises several offices into one schema behind one key. As of November 13, 2026, 15 offices are live with about 64.7 million records: Australia, Brazil, Canada, EUIPO, France, Italy, Mexico, Norway, Singapore, Spain, Sweden, Switzerland, the UK, the US and WIPO Madrid. For EU work that means EUIPO (3,269,808 records, synced daily), France (3,543,314, weekly), Italy (1,413,035, daily), Spain (3,807,383, daily), Sweden (612,529, daily) and WIPO (16,221,392, weekly), plus the UK, Switzerland and Norway, all daily.
Requests go to https://api.signa.so with an Authorization: Bearer $SIGNA_API_KEY header. Keys start with sig_, are self-serve at app.signa.so, and belong on the server only. The parameters below are documented on the search trademarks reference.
Who protects this name in a country?
TMview's territory filter answers "what covers this name in Sweden?", whatever the filing route. In Signa, that is the jurisdictions parameter. This call counts every record with the exact text OATLY whose territory includes Sweden:
curl -sG "https://api.signa.so/v1/trademarks" \
-H "Authorization: Bearer $SIGNA_API_KEY" \
--data-urlencode "mark_text_is=oatly" \
--data-urlencode "jurisdictions=SE" \
--data-urlencode "aggregations=office_code,status_stage" \
--data-urlencode "aggregations_only=true" \
| jq -c '.aggregations'{"office_code":{"EM":15,"WO":7,"SE":2},"status_stage":{"registered":19,"examining":2,"expired":2,"cancelled":1}}Records with the exact text OATLY covering Sweden, by filing route
24 records with the exact text OATLY cover Sweden: 15 EUTMs, 7 Madrid designations and 2 Swedish national records (Signa data, queried November 13, 2026). Only 2 of the 24 records sit in the Swedish register itself; the other 22 are EUTMs and Madrid designations, and the count treats each Madrid designation as its own record. The status split matters too: 19 are registered, 2 are in examination, and 3 are expired or cancelled. Use offices=SE instead and you get only the Swedish office's own register.
Read the records
Switch to q for ranked search and pull the records themselves:
curl -sG "https://api.signa.so/v1/trademarks" \
-H "Authorization: Bearer $SIGNA_API_KEY" \
--data-urlencode "q=oatly" \
--data-urlencode "offices=SE" \
--data-urlencode "limit=3" \
| jq -c '.data[] | {mark_text, tier: .match.tier, office: .office_code, number: .application_number, owner: .owners[0].name, classes: .nice_classes, status: .status.primary}'{"mark_text":"OATLY","tier":"identical","office":"SE","number":"2000-07385","owner":"Oatly AB","classes":[29,30,32],"status":"active"}
{"mark_text":"OATLY","tier":"identical","office":"SE","number":"2009-02610","owner":"Oatly AB","classes":[29,30,32],"status":"inactive"}
{"mark_text":"Oatly iKaffe","tier":"similar","office":"SE","number":"2014-02233","owner":"Oatly AB","classes":[29,32],"status":"active"}The classes are Nice classes, the international system for sorting goods and services into numbered categories. Here, 29 covers foods such as milk substitutes, 30 staples such as coffee and cereals, and 32 non-alcoholic drinks. The 2009 record is inactive, meaning it is no longer in force.
The first row cross-checks cleanly. TMview shows the same Swedish record: SE 2000-07385, Oatly AB, classes 29, 30 and 32, Registered, under ST13 SE500000200007385. Same office data, two front doors.
Catch spellings you didn't type
TMview's interface has a fuzzy search mode, so this is not about TMview lacking one. The difference is a documented, callable response that ranks hits and says why each one matched. Here a variant spelling, otly, runs across four EU markets, live marks only, with phonetic matching turned on (it is off by default, and similarity replaces the default list):
curl -sG "https://api.signa.so/v1/trademarks" \
-H "Authorization: Bearer $SIGNA_API_KEY" \
--data-urlencode "q=otly" \
--data-urlencode "jurisdictions=SE,FR,IT,ES" \
--data-urlencode "status_primary=active,pending" \
--data-urlencode "similarity=identical,fuzzy,phonetic" \
--data-urlencode "limit=6" \
| jq -c '.data[] | {mark_text, tier: .match.tier, via: .match.via, office: .office_code, number: .application_number, owner: .owners[0].name}'{"mark_text":"OTLY","tier":"identical","via":["identical","fuzzy","phonetic"],"office":"EM","number":"016403446","owner":"Otly B.V."}
{"mark_text":"Otely","tier":"near_identical","via":["fuzzy","phonetic"],"office":"FR","number":"5098496","owner":"Orderoom"}
{"mark_text":"Otlo","tier":"near_identical","via":["fuzzy","phonetic"],"office":"EM","number":"018638853","owner":"Norsker Games"}
{"mark_text":"OTY","tier":"near_identical","via":["fuzzy"],"office":"EM","number":"001177468","owner":"OTY S.R.L."}
{"mark_text":"OTL","tier":"near_identical","via":["fuzzy"],"office":"EM","number":"003359528","owner":"Adam Rozanski"}
{"mark_text":"OTY","tier":"near_identical","via":["fuzzy"],"office":"EM","number":"018179737","owner":"Guangdong Yanzhuang Biotechnology Co., Ltd."}match.tier grades the hit and match.via lists the channels that found it: "Otely" in France matched on both spelling and sound, "OTY" on spelling alone. Five of the six are EUTMs, which appear because an EUTM protects all four countries. These are search results, not conflicts.
The same in TypeScript
Install the SDK with npm install @signa-so/sdk:
const signa = new Signa({ api_key: process.env.SIGNA_API_KEY });
// Sound-alikes and typos across four EU markets, live marks only
const page = await signa.trademarks.list({
q: 'otly',
jurisdictions: ['SE', 'FR', 'IT', 'ES'],
status_primary: 'active,pending',
similarity: ['identical', 'fuzzy', 'phonetic'],
limit: 6,
});
for (const hit of page.data) {
console.log(hit.mark_text, hit.match?.tier, hit.match?.via.join('+'), hit.office_code, hit.application_number);
}It returns the same six rows as the curl call.
Limits, stated plainly
- Signa is a third-party index. The offices remain the systems of record.
- Coverage is narrower than TMview's. Signa has 15 offices; TMview has 81. Signa lacks most EU national registers TMview has, such as Germany's DPMA and Benelux's BOIP, and has no China, Japan or Korea. It does have Singapore, which TMview lacks.
- Usage is metered. Searches are billed per page in credits. The free plan has a daily cap of 5,000 credits and per-minute rate limits; see pricing.
- Monitoring (watches, alerts, webhooks) is live in beta.
Which Option to Use
- Occasional checks, or a few names to watch: TMview, free, with saved searches and alerts.
- Official EU register data in code: EUIPO's Trademark search API.
- A market Signa doesn't cover: that office's own service, or TMview by hand.
- Multi-office search with similarity ranking, inside a product, without an approval wait: a trademark API such as Signa. The best trademark API comparison and trademark APIs compared lay out the field.
Many teams combine the two: TMview by hand for a final look across all 81 offices, an API for everything that runs in code. The European trademark search guide and the international trademark search guide cover the manual side.
To try the jurisdiction search above, get an API key on Signa's free plan at app.signa.so (no approval step) and follow the quickstart.
This article is for information only and is not legal advice. Consult a trademark attorney for legal guidance specific to your situation.
