# Arieta Data

> Arieta Data licenses EU legislation, CJEU case law and national legislation as structured AI training, retrieval (RAG) and evaluation data: parsed to article level, versioned by content hash and linked by a typed citation graph. Arieta Data is the dataset business of Arieta Labs Ltd, the company behind LexAPI.

## Key facts (as of the November 2026 snapshot)

| Metric | Value |
| --- | --- |
| Documents | 419,355 |
| Articles (estimated from a sample) | 4,070,553 |
| Document versions | 15,098 |
| Citation edges | 801,149 |
| Approximate tokens (characters ÷ 4) | 2,050,707,780 |
| Languages | 24 |
| Jurisdictions with records | 16 |
| Date range | 1801-06-10 to 2026-11-12 |
| Snapshot | 2026-11 (November 2026) |

Scope of every snapshot: EU legislation, CJEU case law, national legislation and national case law (3,496,146 decisions), the last delivered with personal data masked. UK case law is not offered: the Open Justice Licence does not cover computational analysis. Custom cuts per jurisdiction, language, date range or document type are prepared within 5⁠–⁠6 business days. We reply within one business day.

Documents and citation edges are exact counts. Articles and tokens are estimated from a 2% random sample of rows, extrapolated per lane (small lanes are counted exactly); tokens are characters ÷ 4.

## FAQ

### What is Arieta Data?

Arieta Data is a licensable dataset of European law for AI training, retrieval and evaluation, from Arieta Labs Ltd, the company behind LexAPI. It contains EU legislation, CJEU case law and national legislation from 15 countries, parsed to article level, with version history and a typed citation graph. As of the November 2026 snapshot the corpus holds 3.9M+ documents, of which 410k+ documents (4M+ articles) are ready to license.

### Can I legally train an AI model on EUR-Lex data?

EUR-Lex’s legal notice allows reuse of the legal documents it publishes, including commercially, with acknowledgement of the source; Commission documents follow the Commission’s reuse policy (Decision 2011/833/EU) and CJEU case law the CURIA copyright notice. National sources each have their own terms, and UK case law needs a separate licence for computational analysis. Arieta Data documents the terms per source. See [Is EUR-Lex data licensable for AI training?](https://data.arietalabs.com/guides/eur-lex-data-licence-ai-training/). Not legal advice.

### Is national case law included?

Yes. CJEU case law is included as published by the Court. National case law is included too (3.4M+ decisions as of the November 2026 snapshot): court decisions name or pseudonymise people, so they are delivered with personal data masked, including magistrate and clerk names for France. Tell us which courts you need, subject to licence terms. UK case law is not offered: the Open Justice Licence does not cover computational analysis, and the separate licence from The National Archives cannot be passed on.

### What formats, delivery and updates do you offer?

Apache Parquet (zstd, sharded at about 512 MB) and gzip JSON Lines with the same fields, by signed download or to a bucket you control. Each snapshot ships MANIFEST.json with SHA-256 checksums per shard, a datasheet and LICENCES.md. Updates are monthly deltas with upserts and tombstones.

### How fast can we get the data?

The free sample downloads now, with no sign-up. Custom cuts per jurisdiction, language, date range or document type, and licensed deliveries, are prepared within 5⁠–⁠6 business days. We reply within one business day. See [Request access](https://data.arietalabs.com/request-access/).

### How much does it cost?

Sample records are free to download. Research / Startup licences start at €7,000 a year, Commercial at €18,000 a year, and Foundation / Lab at €50,000 a year (custom terms). Commercial and Foundation / Lab include a perpetual right to the snapshots you received. See [Licensing](https://data.arietalabs.com/licensing/).