Home / Guides / Calculator Free Tiers

Free Duty Calculator Limits, Compared: What Each Tool Allows Before You Pay

Updated October 2026. Every figure below is recomputed at build time from our competitor measurement registry (live_test_oct2026, Playwright headless-Chromium real submissions plus curl/UA page fetches); provenance notes are embedded in the page source.

Every duty calculator advertises something free. Almost none of them advertise what the free thing is worth, and the industry’s dirtiest little trick is hiding a quota behind the word free. Between 2026-10-11 and 2026-10-18 we ran live submissions and page captures against 6 tariff tools — real form posts, not marketing-page readings — and logged exactly what each one allows before asking for payment. The spread is wider than most comparisons admit: one major tool caps anonymous users at a handful of lookups a day, one publishes no allowance at all, and one lets you calculate without even creating an account. This page shows the full census, the importer arithmetic that turns a daily cap into a monthly bill, and a worked case where a “free” calculation returned a number that would have overstated duty by an order of magnitude.

1. The census: six tools, three behaviors

The registry stores one cell per tool, free_tier_daily, and every cell carries a status, a verbatim quotation, a source URL and a capture timestamp. Verified cells were measured directly; unverified ones stay deliberately empty of numbers. Here is the population as measured in the 2026-10-18 window:

Recomputed from data/competitors.json (registry live_test_oct2026; replay with python3 .state/scripts/topic_pool_batch_probe.py, probe key K9). Raw captures and byte fingerprints: .state/media/output/competitor-oct/.

Three behaviors, three buying situations. If you classify five products a week, every tool here is functionally free and you should optimize for answer quality instead (sections 3–4). If you quote dozens of SKUs a day against customer enquiries, a daily cap converts directly into either a subscription or lost quotes — do the multiplication on your own catalog before signing up anywhere. And if you are building an automated pipeline, distinguish a free tier from a free trial: the first is a permanent floor you can architect against, the second is a countdown clock.

2. What a 5-a-day cap means in importer arithmetic

A single number like “5 free calculations per day” hides enormous variance in what it costs different businesses. Work it through: at 5 lookups a day the monthly ceiling is 150 calculations. A sourcing agent triaging three new supplier quotations a day never touches the ceiling. A distributor preparing a 40-line customs pre-classification sheet needs 8 working days at the cap before the sheet is done — and that is before a single retry — so either the deadline moves or the overflow goes onto the paid meter, which at the advertised £0.10 per API call (SimplyDuty bills in pounds sterling) prices out at a full month’s worth of a modest catalog before you reach the second pricing tier. The economics flip again for seasonal buyers: a cap that binds for two weeks of each quarter is invisible the other ten, which is precisely when the upgrade prompt finds you mid-season with quotations pending.

There is also a silent tax on retry behaviour. Real classification work involves mistakes: a typo’d HS input, an ambiguous product description, a re-check after a colleague’s comment. Every retry is another calculation against the cap. On a strict daily quota, the effective throughput is lower than the headline number suggests — plan roughly a third of the nominal allowance for productive work and the rest for corrections, or pick a tool whose pricing model does not charge you for your own proofreading. When the arithmetic gets serious, our landed-cost duty calculator prices the full US stack without a lookup counter, and the specific and compound tariff arithmetic walkthrough shows what the rate column is actually doing underneath any calculator’s UI.

3. Quota is not quality: the wine case

A generous allowance for wrong answers is expensive at any price. During the same window we submitted three scripted scenarios to each calculator that accepts interactive input; here is everything SimplyDuty returned, best case first. The smartphone case (8517.13.00.00, $1,000 declared) came back duty-free at 0%, matching the schedule — a clean hit. The wine case did not go as well: scenario B declares $500 of still wine under HS 2204.21, whose correct treatment is 6.3c/liter specific (2204.21.50 Other); effervescent .20 = 19.8c/L, yet the captured result read “mapped to 2204.21.2000 (effervescent); Duty 51.37 USD at Rate of 57.3% computed on an origin-currency basis - SimplyDuty reads Product value 'In currency of origin', so our 500 entered as intended-USD was treated as CNY; capture prints 'exchange rate: CNY:USD 0.149429' and CIF base (500 goods + 100 shipping) x 0.149429 = 89.66 USD -> 51.37/89.66 = 57.29%”. The third submission then hit the register wall exactly as the cap implies — 3rd anonymous submission printed register wall ('register for 5 free calculations per day') yet still returned Duty 1201.41 USD (Rate of 40%) incl. China add-ons; base 2.5% not shown separately (paywall position: after 2 clean anonymous calcs this session). Anonymous quota in practice: two clean calculations, not five.

Two compounding errors produce the wine number. First, resolution: the tool mapped the input to the effervescent-wines subheading rather than the still-wine line, doubling the specific component. Second, the basis: it treats the product value as being in the currency of origin, so dollars entered as dollars were converted at the exchange rate shown on the capture, turning a cents-per-litre question into a percent-of-a-mis-converted-base calculation. Neither error is exotic; both are exactly the failure modes that surface when a calculator must guess at inputs a licensed broker would confirm. The lesson generalises: before trusting any free calculator’s output on a high-duty line, verify the subheading resolution and the valuation basis against the schedule itself. Our tariff-stacking walkthrough shows how the same verification habit catches Chapter 99 surprises too.

4. Null is not zero: measuring honestly

Most comparison pages paper over the tools they could not test. We do the opposite: 3 of 6 cells in this census are explicitly marked unverified, each with the reason attached. Zonos ships genuinely useful free calculators but publishes no query allowance anywhere on its tools or pricing pages — neither “0” nor “unlimited” fits without inventing a number. Easyship advertises a “Free forever” plan, but its plan ladder renders prices client-side and only the $0.005-per-call overage is statically readable, so the allowance remains unmeasured. ImportYeti’s pricing page has been Cloudflare-gated (HTTP 403) through our egress on the last two measurement rounds. Publishing the hole instead of papering over it is a feature: next month’s refresh either fills it with a number or documents that the wall still stands.

This discipline has a compliance cousin. When you read a competitor’s “free” claim, ask what the denominator is: free calculations (SimplyDuty), free trial (Otterly), free plan with metered overages (Easyship), or free forever with no published limit (ours, sentinel −1 in the registry). Four different offers, four different renewal conversations — and only one of them appears in a registry cell with a verbatim quotation behind it. The machine-readable version of this entire comparison lives on our competitor measurement dashboard, refreshed on the monthly cadence the mission document prescribes.

5. The AI-citation angle: who is even findable

There is a second axis to “free” that the trade-press comparisons miss: whether a tool can be cited by the AI assistants importers increasingly ask first. We snapshot two signals per tool — robots.txt posture toward AI crawlers, and publication of an /llms.txt index:

Cells from data/competitors.json (response_speed_s, llms_txt, ai_crawler_open). Robots posture is scored 0–5, where 5 means a standard open robots.txt with no AI-specific blocks; a dash marks “not measured this round”; speeds are medians of five homepage GETs, all 2026-10-18 unless blank.

All six tools leave their doors open to standard crawlers, so discovery is not the differentiator — citability is. Just 3 of 6 publish an /llms.txt machine index; SimplyDuty’s returns HTTP 404 despite being the category’s best-known brand. When an assistant summarises “the free limits of duty calculators”, sources that expose structured, quotable summaries get cited and sources that hide behind JavaScript pricing ladders get paraphrased at best. Speed compounds it: our homepage answered in 0.19 s median versus 2.0 s for the slowest verified competitor, and response latency correlates with whether a busy model-assisted workflow bothers citing you at all. Being generous for free is worthless if the generosity cannot be found.

6. Playbook: choosing a free calculator without the hangover

1. Read the cap, then multiply by your catalogue. Headline “free” claims without a number (or with the number buried in an FAQ) are the weakest offer in the market; a verified 5/day beats an unverifiable “free forever” for planning purposes.
2. Price the retries. Assume a third of your lookups are corrections; a cap that fits your clean-pass volume may not survive contact with a real working day.
3. Check the basis before the rate. Currency-of-origin confusion (the wine case above) and subheading mis-resolution are the two errors we caught live; both change the answer more than any tier difference between tools.
4. Budget the fees around the duty. Cap or no cap, entry costs continue after the duty line — our MPF and HMF entry fee guide covers the fixed charges no calculator bundles by default.
5. Compare against a registry, not a landing page. Any vendor can rewrite a pricing page overnight; a dated cell with a verbatim quotation and a byte fingerprint cannot quietly move. Hold every tool — including ours — to that standard, and hold our own commercial terms to it too on the pricing page.

7. Frequently asked questions

1. How many free calculations does SimplyDuty allow?

Five per registered account per day. We verified the wording on the live SimplyDuty homepage during the October 2026 measurement window (“you will get 5 free calculations per day”), and its own pricing page repeats the same allowance. At one lookup per SKU that is a hard ceiling of about 150 classifications a month before the paid meter starts.

2. Is Otterly really a zero free tier, not just a small one?

Yes — its own pricing FAQ states there is a free trial but no free tier, which we re-read on the live page in the same window. A trial expires; a free tier does not. The distinction matters when you budget a recurring classification workload.

3. Why are some tools marked unverified instead of zero or unlimited?

Because we could not measure a number, and writing 0 or “unlimited” would both be wrong. Zonos and Easyship publish calculators but no query allowance anywhere on their pages (Easyship’s plan ladder is client-rendered), and ImportYeti’s pricing page served HTTP 403 through our egress on the last two rounds. An unmeasured cell stays null in our registry; it is never quietly converted into a flattering number.

4. Can a free calculation still cost me money?

It can. In our October 2026 live submission, a $500 wine order under HS 2204.21 came back from SimplyDuty as $51.37 — an implied 57.3% rate computed on a currency-converted basis — while the schedule answer is a specific cents-per-liter rate. Resolution errors (picking the effervescent subheading) and basis errors (treating dollar input as origin currency) both flow straight into your deposit at entry.

5. What does ExportSonar itself offer for free?

Instant US tariff estimates by HTS code with no sign-up and no daily lookup cap — the −1 sentinel in our own registry row. The catch we publish alongside it: unlimited lookups are only useful if the answers carry their provenance, which is why every number on this page cites a measurement command or an evidence file.

8. Recap

• Six tools measured (2026-10-11 to 2026-10-18): SimplyDuty 5/day (verified), Otterly no free tier (verified), ExportSonar uncapped no-signup (verified), and 3 tools with genuinely unpublished allowances — recorded as null, never as zero.
• 150 lookups a month is the practical ceiling of the best-known free tier; quota math decides who upgrades far more than the sticker price does.
• Accuracy beats generosity: a live wine shipment priced at an implied 57.3% on a mis-derived $89.66 basis shows why verification habits matter more than daily allowances.
• Citability is the hidden free tier: 3/6 tools publish an AI-readable index, and the ones that do are the ones assistants can recommend.