Guide
The USPTO API, explained
The USPTO publishes free APIs and bulk data — but none of them do full-text search. Here's what each official option covers, and where a dedicated trademark search API fits in.
Last updated July 2026
What the USPTO offers developers
The USPTO's Open Data Portal exposes trademark data three main ways:
TSDR (Trademark Status & Document Retrieval) — fetch the status, prosecution history, and documents for a single application by serial or registration number. Free with an API key, with conservative rate limits. Ideal when you already know exactly which mark you're looking at.
Trademark assignment search — ownership transfer records, again keyed to known marks.
Bulk data downloads — daily and annual XML files of the entire register. Complete, but you're signing up to parse gigabytes of XML, keep it current every day, and build your own search index on top.
The gap: search
None of the official endpoints answer the question most products actually have: “what marks look like this one?” There is no official API for fuzzy name search, autocomplete, owner portfolio lookups, or similarity screening. The USPTO's own TESS search is a website, not an API.
That's the layer Markbase provides. We ingest the USPTO bulk files daily into a search index of 14M+ marks, then expose it as a REST API: typo-tolerant search, autocomplete, owner lookups, clearance screening, and aggregate stats — with the raw record data (serial number, status, classes, dates, owner, goods & services) in every response.
Side by side
| Task | Official USPTO APIs | Markbase |
|---|---|---|
| Lookup by serial number | TSDR | GET /trademark/{serial} |
| Search by mark name | — (website only) | GET /search — fuzzy, typo-tolerant |
| Autocomplete | — | GET /suggest |
| Owner portfolio | — | GET /owner |
| Similarity screening | — | GET /clearance |
| Full register access | Bulk XML (DIY parsing) | Indexed daily, queryable instantly |
If your product only ever needs the status of marks it already knows about, TSDR is free and official — use it. The moment you need search, screening, or anything real-time in front of users, building on the bulk files yourself is weeks of work that a search API replaces with one HTTP call.
Try the same query both ways
Finding registered marks named “acme” via Markbase:
curl "https://api.markbase.co/search?q=acme&status_code=800"
With the official APIs, the equivalent requires downloading the bulk XML, parsing it, and building an index first — there is no request that does this directly.
Start with a free API key
Try any endpoint from curl right now — then sign up free for a key, higher limits, and usage analytics.
Frequently asked questions
Does the USPTO have an official API?
Yes. The USPTO publishes several APIs through its Open Data Portal, including TSDR for trademark status and documents, an assignment search API, and downloadable bulk data files. They are free but built around exact serial or registration number lookups rather than search.
Can I search trademarks by name with the official USPTO API?
Not really. TSDR retrieves one record at a time by serial or registration number. To search by mark name, owner, or class, you either parse the bulk XML files yourself or use a search API built on top of that data, which is what Markbase is.
Is Markbase affiliated with the USPTO?
No. Markbase is an independent product of Rad Soft that indexes the USPTO's public bulk trademark data daily. It is not endorsed by or affiliated with the USPTO.
How fresh is Markbase data compared to the USPTO?
Markbase ingests the USPTO's daily bulk files every day, so records are typically within 24 hours of what the USPTO publishes.