Export Classification

Export classification 

Export control classification codes determine whether a product is subject to export controls, licensing requirements, or restrictions when it leaves a country. In the United States, these are Export Control Classification Numbers (ECCNs) — for example, EAR99 for most general consumer goods, or controlled codes such as 3A001 for certain electronics. Other export schedules use their own equivalents.

Assigning the correct export control code is essential for trade compliance. The right code drives licensing decisions, screening against restricted-party and destination rules, and accurate documentation. Missing or incorrect codes can lead to held shipments, penalties, and compliance risk.

Zonos Export Classification predicts the most likely export control code for a product so you can fill gaps in your catalog and support your export compliance review.

How it works 

Zonos Export Classification uses AI to analyze the product attributes you provide — including the name, description, categories, material, and brand — together with the ship-from country, which determines the applicable export schedule. From these inputs it predicts the most likely export control code and returns it with a confidence score and a ranked list of alternates.

You may optionally provide an export HS code (such as a Schedule B code) for the product. If you don't, Zonos determines one automatically as part of the classification.

Prediction quality correlates directly with input quality. Detailed, accurate product information produces higher-confidence results, while sparse or generic descriptions provide less to work with.

Confidence scoring and alternates 

Every prediction includes a confidenceScore between 0.0 and 1.0 for the primary code, where higher values indicate greater certainty. Because some products map cleanly to one code while others are genuinely ambiguous, each result also includes a ranked list of alternates, each with a probabilityMass reflecting its relative likelihood.

We recommend establishing confidence thresholds that match your compliance requirements — for example, accepting high-confidence primary predictions automatically while routing lower-confidence results to manual review, where the alternates can help guide a decision.

Infer an export classification 

Zonos Export Classification is currently available via API only.

To infer an export classification, use the exportClassificationsInfer mutation. The name and shipFromCountry fields are required; providing more detail — brand, categories, description, and material — improves prediction accuracy and confidence. You can also provide an exportHsCode (such as a Schedule B code); doing so is recommended for the most reliable results, and if it is omitted Zonos determines one for you. You can submit multiple items in a single request.

In the example above, the product is predicted as EAR99 with very high confidence, meaning it is not specifically controlled. The alternates show other codes the model considered, which you can use to evaluate edge cases when the primary prediction has lower confidence.

Implementation 

We recommend deploying Zonos Export Classification with confidence thresholds aligned to your compliance standards and review workflows for low-confidence results. Predictions are meant to enhance — not replace — your existing export compliance processes by filling information gaps and highlighting where additional review is warranted.

For broader product classification, including HS codes used for landed cost and customs, see Zonos Classify.