You are three weeks into a project that has to pull train and railcar data into your company's logistics tools. Your manager asked a simple question: which railroads publish APIs worth building on? So you open the Rail industry page on apis.io, the way anyone would, and read the table from the top.
The page describes itself plainly. It covers companies operating freight railroad networks that move agricultural products, energy, industrial goods and intermodal containers. It holds 30 providers and 149 APIs. You scan the leaders and something is off.
The highest-scoring provider is American Express Global Business Travel, at 49.8 on the Kin Score, in the developing band. Second is the U.S. Department of Transportation at 49.5. Neither one runs a single train. The first company that actually does is BNSF, in third place at 47.6. After that comes the OpenTravel Alliance at 45.7, then Cedarai at 39.1 and TransportAPI at 34.8, both in the thin band. Nobody in rail reaches strong.
You wonder how a freight vertical ends up led by a travel agency and a federal department. The post gives the reason: membership is computed from tags, not from a fixed list. Those tags pull in travel distribution, which brings Amex GBT and OpenTravel. They pull in a UK transit data service, TransportAPI. They even pull in BT, a telecom carrier, which sits at 8.7. So the top of the table reflects tagging, not track.
You set the leaders aside and look at the railroads themselves. BNSF is the exception. It publishes a real customer API program. One small wrinkle: the industry page still shows its earlier 47.6, while its provider record has since moved to 49.2.
The other big freight carriers sit far below it. Norfolk Southern scores 22.0. Union Pacific scores 20.0, with 8 APIs on record. Canadian National Railway is at 9.2 and CSX at 6.7. Kansas City Southern shows 0.0, and the record describes it as a holding company in the past tense. The gap between BNSF and Union Pacific is 27.6 points, on two networks that serve overlapping shippers. Passenger operators do no better: VIA Rail at 22.9, Amtrak at 18.7 and Deutsche Bahn at 5.7.
Then the picture shifts. The software is on the supply side. The companies selling to railroads publish more than many railroads do. Cedarai's ARMS External API exposes REST and gRPC services for railcar inventory and history, with 9 APIs on record. Railinc, the Association of American Railroads subsidiary, scores 13.0 with 3 APIs. Equipment makers trail: Greenbrier at 25.8, Wabtec at 17.9 and Trinity Industries at 3.4.
Regulators show up on both sides of the line. DOT is second overall with 41 APIs, yet the Federal Railroad Administration, the agency that actually regulates rail, scores 18.2.
This is where the post turns on its own catalog. It does not just blame the railroads for being quiet. It admits that the ranking you first read is an artifact of how apis.io builds the list. A freight vertical led by companies that do not move freight is the catalog's own doing, not a finding about the industry. The post says so directly and tells you to read the table with that in mind.
So you change your plan. You stop treating the top row as a recommendation. For carrier data in North America, you start with the BNSF Trains API, the one carrier with a contract-backed surface you can build on today. For railcar data across carriers, you look at Cedarai and Railinc before the railroads.
You take away two things. BNSF is the carrier to build on, and the suppliers are where cross-carrier railcar data lives. And a ranking is only as good as the tags behind it: the top two spots belong to a travel agency and a regulator because of tags, not track.
Read the original: https://apis.io/2026/11/14/rail-on-apis-io-is-a-freight-vertical-led-by-a-travel-agency-and-a-regulator/