Codelob - 2026 Pricing, Features, Reviews & Alternatives

Three questions nobody else answers in one place

Reference data already exists — scattered across portals, blind to versions, and full of links that should never have been drawn. That is the gap.

- Country × version × date: Ask which NACE is in force in Türkiye today and you get the national derivative, not the European root. National versions are separate versions of the same classification, resolved by country, version or date — never merged into one ambiguous list.

- Pin a version, stop the breakage: Pin an exact version and the answer never moves under you. Superseded versions stay online — NACE Rev. 2 is still queryable next to Rev. 2.1 — so a code written on a 2019 document still resolves. A since parameter returns only what changed between two versions.

- Crosswalks that state their trust: Every link carries a trust level and the sentence that justifies it. Links that must never be drawn — occupation to industry — are not weak entries in the data; they do not exist at all, and the loader refuses to write them.

Built for the teams a wrong code costs the most

- Master-data teams: Version transitions with evidence of what was in force when, and a country axis that nobody has to maintain by hand.

- Health, HR and commerce software: ICD-10, ISCO, ESCO, NACE and CPA already loaded, already labelled, already linked — instead of a spreadsheet somebody imported years ago.

- Developers and startups: JSON over HTTP, no key on the open core, a delta feed for the mirror and a fingerprint to test against.

- Data and AI pipelines: Official crosswalks with their trust level, alternative labels for matching, and every row recomputable from its own fingerprint.