AI HTS code classifier
Describe your product in plain English and get the most likely HTS codes — with GRI reasoning and real duty rates. With ~19,000 tariff lines, the wrong code can cost 10–50+ points in duty. Free.
How it works
Describe it
Type what your product is, what it's made of, and its use — plain English is fine.
AI applies the rules
A model suggests the most likely HS subheadings using the General Rules of Interpretation (GRI), with its reasoning.
Grounded & verified
Each suggestion is matched to real HTS lines with live duty rates, linked to landed cost. Confirm the final code before you file.
Frequently asked questions
How accurate is the AI HTS classifier?
It is decision support, not a binding ruling. An AI model suggests the most likely 6-digit HS subheadings using the General Rules of Interpretation, and ShipSeeker grounds each suggestion in the real HTS database (actual tariff lines and rates). The importer of record is legally responsible under the "reasonable care" standard — confirm the final 10-digit code against CBP CROSS rulings or request a binding ruling.
Why does classification matter so much?
The U.S. HTS has ~19,000 lines, and with 2025–26 tariff layers the duty difference between two plausible codes can be 10–50+ percentage points. Misclassification can cause overpayment, or underpayment that triggers CBP penalties and retroactive duty bills.
What is a binding ruling?
If you need certainty, you can request a binding classification ruling from CBP (eRulings). Existing rulings are searchable on CBP CROSS — this tool links your product term straight into a CROSS search.
AI-assisted decision support — not a binding ruling or legal advice. The importer of record is responsible for the correct 10-digit classification under CBP's "reasonable care" standard. Suggestions come from an AI model applying the GRI, grounded in the real HTS database; always confirm the final code against the USITC HTS and CBP CROSS rulings, or request a binding ruling for certainty.