Purchasing Guides / Tariff Code Classification

Tariff Code Classification

A purchasing guide to buying a workflow that suggests import tariff codes for your products, with the tariff text and past customs rulings behind each suggestion, for your trade compliance team to approve.

Yearly running cost

Assuming 2,400 products a year

By hand
$27K / year
600 hours of work
Subscription
$4.5K+ / year
This workflow
$112 / year
Machine usage only; setup, hosting and review are extra.
What should it classify?

Suggest a tariff code for each product with the reasoning, tariff text and rulings behind it. Your trade compliance team approves the code.

One product includes up to 5 pages of product description and specification.

Your trade compliance team or customs broker approves every code. No customs filings or binding ruling requests.

What is included

Read product descriptions and specification sheets, search the tariff schedule and published rulings, and suggest a code with step-by-step reasoning, cited text and any missing product facts.

Start with clear product descriptions, materials and uses, and codes your customs broker has already approved. The catalog option reads your product records and rechecks codes when the tariff schedule changes; it does not file customs entries.

How accurate should the suggested codes be?

A code can look right at a glance and still be wrong in the final digits. Choose how often suggestions must be right at each level and which uncertain products must be flagged.

How these standards are measured

Test on held-out products classified by a licensed customs broker or trade specialist, from product families not used for setup. Measure the internationally shared code level and the full national code separately. Products that need more facts or a ruling must be flagged, not guessed.

Targets for your selected standard
What is checkedTarget
Shared code level correctSuggestions matching the specialist’s code at the internationally shared level, divided by all test products; flagged products count as not correct.≥95%
Full code correctSuggestions matching the specialist’s full national code, divided by all test products; flagged products count as not correct.≥85%
Uncertain products flaggedProducts the specialist labelled as needing more facts, a ruling or expert judgement that were flagged, divided by all such labelled products.≥95%
Valid citationsEvery cited heading, note and ruling must exist in the stored tariff edition or ruling database and say what the reasoning claims.≥100%

The importer remains responsible for classification. A suggestion is research for your specialist, not a ruling; every cited heading, note and ruling must exist in the current tariff edition or ruling database.

How quickly do you need a suggestion?

Choose how quickly a suggested code should be ready after a product is submitted. Your specialist’s review time is separate.

Timing details

Time from a complete product submission to a stored suggestion, including reading specifications, searching tariff text and rulings, queueing and retries. Confirm document sizes and peak volume with your provider.

The target applies to at least 95% of agreed test runs, with 4 in progress at a time.

How much do you want to spend per product?

Choose the machine budget for researching each product. Tighter budgets may search fewer rulings or use smaller models.

Cost details

Includes model calls, retries and shared hosting allocation. Specialist review, customs broker fees, duties and the calling agent are separate. Failed and repeated runs also consume resources.

Reference machine cost: $0.15 – 0.35 per product at 200 products a month. The selected cap is a target to test, not a replacement for this estimate.

Where do you want it to run?

Keep product data in your cloud or on your own server. Choose whether product descriptions may go to an approved AI service.

Data and access details

Runs in a cloud account you control, with access controls and logs.

Only approved product description fields go to the selected external model. Agree access and retention first.

The public tariff schedule and published rulings are stored with the workflow and updated on a schedule you approve. No external calls uses private inference; supplier names and costs never need to leave your environment.

How do you want to use it?

Use a review page, your current AI agent or a dedicated classification agent. You can choose more than one.

Anything else your provider should know?

Optional. Your choices are included automatically.

Common questions

Do I need this if my customs broker already classifies my products?

Not necessarily. If your product range is small and stable, your broker may be enough. Buy this when new products arrive often, research takes your team or broker significant time, or you want consistent evidence for every code.

Does an approved suggestion protect me in a customs audit?

No. The importer remains responsible. The workflow gives your specialist reasoning and citations to review; it is not a binding ruling from customs authorities.

Does it calculate duties or extra tariffs?

No. It suggests the code and shows its evidence. Duty rates, additional tariff measures and exclusions should be checked in your customs process.

Which countries does it cover?

The guide covers United States import codes. Other countries share the international code level but differ in their national digits, so each additional schedule is extra scope.

Can I run it without an external AI service?

Yes, with private inference on hardware you supply. Measure reasoning quality on your difficult products and running cost during the pilot.

Download RenX

Get the app.

Install, sign up, and start on your free plan with welcome credit included. No credit card required.

On a platform not listed? Leave your email and we'll notify you when a build is available.