German Federal Court Decisions – BGH, BVerfG, BAG & more to JSON / Markdown

Get decisions of Germany's federal courts – Bundesgerichtshof (BGH), Bundesverfassungsgericht (BVerfG), Bundesverwaltungsgericht (BVerwG), Bundesfinanzhof (BFH), Bundesarbeitsgericht (BAG), Bundessozialgericht (BSG), Bundespatentgericht (BPatG) and the Joint Senate (GmS-OGB) – as clean, structured JSON with Markdown or plain text. The data comes from the official open-data portal rechtsprechung-im-internet.de run by the Federal Ministry of Justice (BMJV), which lists more than 84,000 decisions since 2010.

$1 per 1,000 decisions. You only pay for decisions returned. Downloads that fail are listed with status: "error" and are free.

Why this actor

• ✅ Official source, no scraping of search pages: reads the ministry's own table of contents (rii-toc.xml) and the official decision XML files
• ✅ Structured sections: leitsatz (headnote), tenor (operative part), tatbestand, entscheidungsgruende, gruende, abweichendeMeinung, cited normen, vorinstanz (lower instances), ecli
• ✅ LLM-ready text: the whole decision as one Markdown document with headings and margin numbers (Rn.) kept, or as plain text
• ✅ Filters: court, date range or "last N days", file number (Aktenzeichen), judgment vs. order, keywords
• ✅ Fast and polite: the table of contents is downloaded once per run, at most 4 parallel downloads. 3 decisions finish in about 15 seconds
• ✅ Public domain data: court decisions are official works under § 5 UrhG (German Copyright Act) and are not protected by copyright. Decisions on the portal are already anonymised by the courts

Who it's for: legal-tech teams, law firms, in-house legal departments, researchers and journalists – and anyone building RAG pipelines or AI legal assistants on German case law.

Use cases

• Legal research: pull every BGH tenancy-law decision (fileNumber: "VIII ZR") of the last year in one dataset
• RAG for legal assistants: feed fullTextMarkdown with ECLI and Aktenzeichen as metadata into your vector database (Pinecone, Qdrant, pgvector, …)
• Monitoring new rulings on a topic: schedule a daily run with lastDays: 7 and keywords: ["DSGVO"] and get notified (e-mail, Slack, webhook) when the federal courts publish something new
• Datasets and analytics: statistics on courts, senates, document types and cited norms

Input

Field	Default	Description
courts	["BGH"]	Any of BGH, BVerfG, BVerwG, BFH, BAG, BSG, BPatG, GmS-OGB. Empty = all courts
lastDays	30	Only decisions dated within the last N days (ignored if dateFrom is set, 0 = no limit)
dateFrom / dateTo	–	Decision date range, YYYY-MM-DD
fileNumber	–	Aktenzeichen contains, e.g. VIII ZR, 1 BvR, IX ZB 72/08
keywords	–	Match in title, Leitsatz, Tenor, norms (case-insensitive, any keyword)
keywordsInFullText	false	Also search the reasons (Gründe)
documentType	any	Urteil (judgment), Beschluss (order) or any
maxDecisions	3	Maximum decisions returned, newest first
includeFullText	true	Adds Tatbestand, Gründe and fullTextMarkdown
outputFormat	markdown	markdown or plain
maxScan	500	With keyword / type filters: max decisions downloaded to find matches

{
  "courts": ["BGH"],
  "dateFrom": "2026-01-01",
  "fileNumber": "VIII ZR",
  "documentType": "Urteil",
  "keywords": ["Mietpreisbremse", "Eigenbedarf"],
  "maxDecisions": 50,
  "outputFormat": "markdown"
}

Output

One row per decision (real example from a test run, text shortened):

{
  "status": "ok",
  "court": "BGH",
  "senate": "3. Zivilsenat",
  "decisionDate": "2026-10-22",
  "fileNumber": "III ZR 3/26",
  "ecli": "ECLI:DE:BGH:2026:221026UIIIZR3.26.0",
  "documentType": "Urteil",
  "title": "Gemeindliches Vorkaufsrecht, Beginn der Ausübungsfrist; Entstehung",
  "normen": ["§ 24 Abs 1 S 1 Nr 1 BauGB", "§ 28 Abs 1 S 1 BauGB", "§ 28 Abs 2 S 1 BauGB", "..."],
  "vorinstanz": ["vorgehend Hanseatisches Oberlandesgericht Hamburg, 19. Dezember 2025, Az: 1 U 1/24 Baul", "vorgehend LG Hamburg, 22. November 2024, Az: 351 O 2/23"],
  "leitsatz": "Gemeindliches Vorkaufsrecht, Beginn der Ausübungsfrist; Entstehung\n\n1a. Die Mitteilung des Inhalts des Kaufvertrags, mit der nach § 28 Abs. 2 Satz 1 BauGB ...",
  "tenor": "Die Revision des Beteiligten zu 1 gegen das Urteil des Hanseatischen Oberlandesgerichts - Senat für Baulandsachen - vom 19. Dezember 2025 wird zurückgewiesen. ...",
  "tatbestand": "**1** Die Beteiligten streiten um die Rechtmäßigkeit der Ausübung eines gemeindlichen Vorkaufsrechts durch die Beteiligte zu 3. ...",
  "entscheidungsgruende": "**9** Die Revision hat keinen Erfolg.\n\nI.\n\n**10** Das Berufungsgericht hat ...",
  "gruende": null,
  "fullTextMarkdown": "# BGH 3. Zivilsenat, Urteil vom 2026-10-22 – III ZR 3/26\n\nECLI: ...\n\n## Leitsatz\n\n...",
  "textFormat": "markdown",
  "wordCount": 7176,
  "docId": "KORE300832026",
  "sourceUrl": "https://www.rechtsprechung-im-internet.de/jportal/?quelle=jlink&docid=KORE300832026&psml=bsjrsprod.psml&max=true",
  "zipUrl": "https://www.rechtsprechung-im-internet.de/jportal/docs/bsjrs/jb-KORE300832026.zip",
  "lastModified": "2026-11-03T22:09:08.803Z"
}

Margin numbers (Randnummern) are kept as **12** at the start of each paragraph in Markdown (12 in plain text), so you can cite "Rn. 12".

Pricing

Pay per event: $0.001 per decision returned ($1 per 1,000). Failed downloads and decisions removed by your filters are free. Platform compute for a typical run is a fraction of a cent.

API, n8n, MCP

API (Python):

from apify_client import ApifyClient
client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("fast_data_asap/german-federal-court-decisions").call(run_input={
    "courts": ["BAG"], "lastDays": 90, "documentType": "Urteil", "maxDecisions": 20})
for d in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(d["decisionDate"], d["fileNumber"], d["title"])

n8n / Make / Zapier: use the Apify node with actor fast_data_asap/german-federal-court-decisions and "Run actor and get dataset".

MCP (Claude, Cursor, ChatGPT agents): add https://mcp.apify.com?tools=fast_data_asap/german-federal-court-decisions as an MCP server, then ask e.g. "Find the newest BGH judgments on Eigenbedarf and summarise the Leitsätze."

Notes and limits

• Coverage is what the federal courts publish on rechtsprechung-im-internet.de (mainly decisions since 2010). Not every decision of a court is published there.
• Dates filter on the decision date; decisions often appear on the portal a few days or weeks after they were issued.
• Keyword filters need each candidate decision to be downloaded; narrow the date range or raise maxScan for rare terms.
• Only official public open data is used. No personal data beyond what the courts publish in anonymised form.

Deutsche Kurzfassung

Dieser Actor lädt Entscheidungen der Bundesgerichte (BGH, BVerfG, BVerwG, BFH, BAG, BSG, BPatG, GmS-OGB) aus dem offiziellen Open-Data-Portal rechtsprechung-im-internet.de des Bundesministeriums der Justiz und liefert sie als JSON mit Markdown oder Klartext: Gericht, Senat, Datum, Aktenzeichen, ECLI, Dokumenttyp, Titel, Normen, Vorinstanz, Leitsatz, Tenor, Tatbestand, Entscheidungsgründe / Gründe und den Volltext mit Randnummern. Filter nach Gericht, Zeitraum, Aktenzeichen, Urteil/Beschluss und Stichwörtern. Gerichtsentscheidungen sind amtliche Werke nach § 5 UrhG und gemeinfrei; die Veröffentlichungen sind von den Gerichten anonymisiert. Ideal für Recherche, juristische KI-Assistenten (RAG) und die Überwachung neuer Rechtsprechung zu einem Thema. Preis: 0,001 $ pro Entscheidung.