The OpenAPI Standard

About this paper

Every score API Evangelist publishes rests on reading contracts, and almost all of those contracts are OpenAPI. So this report is partly an audit of API Evangelist’s own foundation, and it is written that way.

The evidence is 14,257 OpenAPI documents from 6,750 organizations, containing 458,875 operations, analysed exactly as their authors published them — the largest independent corpus of real-world OpenAPI known to exist, and nearly all of it invisible to surveys that ask developers what they use rather than reading what they shipped. Adoption is 26.2% of the providers tracked in the catalog: a triumph over every competing format and a much smaller number than the industry implies. 32.8% of published documents carry no absolute base URL — the {apiRoot} problem once reported as CAMARA’s turns out to belong to OpenAPI everywhere. OAS 3.2.0, released September 2025, appears in four documents. Swagger 2.0, superseded nine years ago, is still 9.0%. There is no conformance programme at all, and 0.4% of the corpus does not parse. The full report has the version reality, the semantic audit, the coalition, the governance roster whose terms expired three years ago, and a scoring caveat aimed inward — including why the OpenAPI Initiative scores 19.4 on the Kin Score rubric and why that number means nothing.

What's inside

1. Executive summary — four numbers
2. What OpenAPI is, and what it is not
3. The origin — a donation, and the thing it replaced
4. What the body actually publishes
5. The maturity ladder — and why this standard does not have one
6. Version reality — what people are actually writing
7. The coalition
8. The adoption gap — the heart of the report
9. Conformance — there isn't any
10. The regulatory layer — and the standards that carry OpenAPI
11. The scoring caveat — required, and pointed at myself
12. What would have to be true
13. The investable thesis
14. Where this is going

What you get for $500.00