Know what ocean freight actually costs before you book a container

Benchmark rates by lane, with the date they were taken

An ocean freight benchmark is only meaningful with its date attached, because container rates in 2026 have moved faster than quarterly reporting can track. Public indices measure different things: Drewry's World Container Index and Freightos' FBX both publish spot composites, Xeneta publishes contracted as well as spot data, and none of them is your quote.

Read a published index as context, not as a price:

• Freight rate indices are composites across many shippers and volumes. A small shipper's quote sits above the index more often than below it.
• The common indices publish 40ft container rates. No comparable public index exists for 20ft containers, so a 20ft "index rate" you are shown was derived by someone, and the derivation is the thing to ask about.
• Surcharges usually sit outside the index number. Terminal handling, bunker adjustment and peak-season surcharges land on the invoice.

Ocean freight benchmarks disagree by more than most quotes do, and the disagreement is measurable rather than rhetorical. On one trade, within nine days, three publishers gave three numbers:

Ocean freight rates read from the wrong row are the most common way a shipper arrives at the wrong expectation. The three lane and trade readings span $1,189 per container, 20.2% between the highest and the lowest, and the composite in the fourth row sits 26% below the Shanghai to Los Angeles assessment because it averages cheaper trades into the same number. A composite is not a lane rate, and a lane rate is not your quote. A freight forwarder quoting from any of the three owes you the row the figure came from.

SeaRates publishes ocean freight medians by lane with p25 and p75 rather than an average, because averages in this market are dragged by a few large contracted volumes. Where the carrier rather than the rate is the question, capacity and on-time performance by carrier is measured separately: MSC holds 21.4% of world TEU, and Maersk led the top 13 on schedule reliability at 78.2% in May 2026. Every table carries the date of the cut and the number of quotes behind it. Rates are refreshed on a 30-day cycle, and any table older than its cycle is marked rather than quietly left in place — what changed this month carries each refresh with its sample and cut-off date.

Reading the right row still leaves the larger question open, and no freight rate index touches it: how much of your bill the ocean rate even is. SeaRates measured that across the 299 pages ranking for landed-cost queries on 07/26/2026 — 5 of them state a share at all, and the five disagree.

A freight rate quoted 20% apart between two indices moves the delivered total by roughly 14% on a full container and by 7% on consolidated cargo. That is why an index argument is worth less than it looks, and why the delivered total is itemised line by line rather than estimated from the freight.

What we measured across this market, and where the corpus is thin

Ocean freight publishing is easy to criticise and hard to quantify, so SeaRates counted it. On 07/25 and 07/26/2026 we read the full text of 2,001 pages that Google United States ranks in the top ten or cites in AI Overviews across 22 question clusters, from container cost to demurrage to license checks. Pages under 200 words and pages blocked to the crawler were excluded before counting, and a page ranking in two clusters is counted in both.

Freight forwarder content is not thin on numbers; it is thin on the sentence that makes a number checkable. The pattern repeats in every cluster: a price without its date, a ranking without its criteria, a regulator named without the registry behind it. Across the 62 pages ranking for freight forwarder comparison queries, 4 date a price, 3 publish a method, 2 cite a primary source, 1 uses the word median and none states a sample size.

Ocean freight coverage is also thin in a second sense, and SeaRates says so rather than filling the gap. No quote array exists yet for vehicle shipping, so no vehicle median appears anywhere on this site. No permanent lane panel exists yet, so rate material here cites dated indices instead of presenting an average of our own as a market rate. Both gaps are named on the pages where a reader would look for the missing number, because a comparison site that hides what it has not measured is doing the thing it criticises.

Three segments, and why we keep them apart

Ocean freight, car shipping and household moves are three different purchases with three different failure modes, and merging them into one ranking helps nobody. A business importing pallets needs customs capability and lane coverage. A person shipping a car needs RoRo or container expertise and export compliance. A family relocating needs origin packing, destination customs and a survey.

• Freight forwarders — commercial cargo, FCL and LCL, customs brokerage, lane coverage.
• Car shipping — RoRo and container methods, export documentation, destination import rules: what sets the bill for shipping a vehicle overseas.
• International movers — household goods, survey and packing, two regulators across the two legs: the overseas moving section.

Ocean freight, vehicles and household goods were measured separately for the same reason they are ranked separately, and each corpus fails its readers in its own way. In the car shipping corpus, 41 of 47 pages name no regulator at all. In the moving corpus, 12 of 67 name a destination-side cost, 17.9% of the pages measured. Among the 27 mover ranking pages, 11 name a USDOT number against 5 naming the FMC. Three markets, three different blind spots, and a single merged ranking would see none of them.

The same shipping company can appear in more than one segment, and its score can differ between them, because the evidence differs. A freight forwarder with strong China-to-US lane data has told us nothing about its household moving service. Readers unsure which of the three they are buying should start from what "international shipping company" actually means.

Each segment carries its own list, and each list names the metric it was ordered on before it names a company: every ranking on this site, and what each one measures is the index of the eight, with the publisher and reading date on each.

How to tell a verifiable review from a manufactured one

Shipping company reviews are worth reading and worth discounting, and the difference comes down to what a reviewer can prove. Reddit holds a top-10 position for 47 % of the queries in this category, which is a useful signal in itself: shippers trust a forum thread with specifics over a five-star average.

• Check the detail. An ocean freight review naming the lane, the container size, the month and the surcharge that surprised them is describing a real shipment. A review praising "great service" is describing nothing.
• Check the distribution. Genuine review sets have a spread. A wall of five-star ratings posted within a short window is a campaign.
• Check the entity. Compare review dates against the company's registration date. Reviews predating the legal entity belong to a predecessor, and the predecessor's record is the one worth finding.

Shipping company review platforms are also businesses. Profile features, response tools and badges are sold to the companies being reviewed, which does not make the reviews false but does explain why almost every profile converges on the same rating band.

Ocean freight review sets are also smaller than their averages suggest. Of the 26 machine-readable ratings SeaRates found in the company-check cluster on 07/26/2026, twelve rest on fewer than 100 reviews, and across the five companies rated on two or more platforms the median gap between a company's highest and lowest score was 1.9 points out of 5. A rating carried from one platform to another is not the same claim it was on the platform that produced it.

Shipping company reviews in this category sit on a corpus that can be counted: SeaRates captured the 177 pages ranking across the 49 queries on 07/26/2026, spread over 127 domains, 102 of which hold a single URL. Of those, 17 pages sit on platforms that sell the companies they list a profile, a badge or accreditation — Clutch, GoodFirms, the Better Business Bureau, Trustpilot, G2 and Yelp — and 12 pages of them are ranked lists of the kind a shipper reads as a recommendation. Just 3 pages are a register lookup where the licence behind a name can be checked: the FMC's two OTI searches and the FMCSA's SAFER snapshot. Classified by domain by SeaRates on 07/28/2026. For every page in this category that lets a reader verify a company, four rank it on a platform that sells that company its own profile.

How this site is checked, and what stands behind the figures

SeaRates publishes no named author at present, and says so rather than inventing a byline. What stands in place of a name is the method: every figure carries the unit, the date it was read and the corpus behind it, and every count on this site can be repeated against the same search results by anyone who wants to check it.

Methodology is deliberately boring and deliberately specific: credentials are re-checked against the FMC OTI list on a fixed cycle, rate tables carry the date of their cut and the sample behind them, corrections are published rather than quietly edited, and the address for correcting a figure is on every page. A comparison site that cannot say when it last checked something is asking for trust it has not earned.

Ocean freight figures on this site carry three things without exception: the unit, the date they were read, and the corpus behind them. The measurement programme behind the tables above covered 22 question clusters and 2,001 pages on 07/25 and 07/26/2026, at a 200-word minimum, and it is repeatable by anyone willing to read the same search results — which is the only sense in which a number here is independent. Corrections are published with the date they were made rather than applied silently, and the address for sending one sits on every page carrying figures.