You have the data expertise. We have the infrastructure. Package your knowledge into a CorpusAI corpus and earn 80% of every subscription.
Apply to Creator Program ↓Everyone has general AI. ChatGPT doesn't know Norwegian maritime regulations, EU tax compliance details, or your industry's standards. Curated, domain-specific data is what businesses pay for.
80%
Revenue to you
CorpusAI takes 20%. You keep the rest. Monthly payouts.
0 kr
Infrastructure cost
We host the vector DB, embedding pipeline, and AI chat. You just bring the data.
B2B
Ready-made audience
Nordic SMBs already on the platform, looking for domain-specific corpora to subscribe to.
Your corpus price
299 NOK/mo
Subscribers
50
Your monthly revenue
11,960 NOK
299 × 50 × 80% = 11,960 NOK/mo
You Bring the Data
Building a domain-specific AI product normally means setting up vector databases, embedding pipelines, chat interfaces, billing systems, and GPU infrastructure. That's months of engineering work before you earn a single krone.
With CorpusAI, you focus on what you're good at — curating domain data — and we provide the entire platform. Upload documents, set your price, start earning.
Four steps from domain expert to earning creator.
Tell us about your domain expertise and the corpus you want to build. Our team reviews applications within 48 hours.
Upload documents, configure automated scrapers, or use our API. We handle chunking, embedding, and vector indexing automatically.
Choose your subscription price (suggested range: 99–999 NOK/mo). Write a description. We handle Stripe billing and invoicing.
80/20 split. You earn 80% of every subscription. Monthly payouts. Transparent dashboard showing subscribers, queries, and revenue.
Your corpus, accessible from anywhere. Upload via API, query via MCP, build integrations that keep your knowledge current.
Push documents programmatically. Every upload gets chunked, embedded, and indexed — ready for RAG queries within seconds. Attach rich metadata so your AI can filter and cite precisely.
Upload a document
curl -X POST \
-H "X-API-Key: YOUR_API_KEY" \
-F "file=@contract.pdf" \
-F "title=Service Agreement Q1 2026" \
-F "category=contracts" \
-F "author=Legal Team" \
-F "tags=compliance,gdpr,saas" \
-F "source_url=https://internal.example.com/docs/123" \
-F "publication_date=2026-01-15" \
https://ai.bluenotelogic.com/api/v2/documents.phpResponse
{
"success": true,
"message": "Document uploaded and indexed (14 chunks, 4,230 words)",
"document": {
"id": 42,
"title": "Service Agreement Q1 2026",
"chunk_count": 14,
"word_count": 4230,
"status": "ready"
}
}POST
Upload & index
GET
List documents
DELETE
Remove & clean up
• Auth: X-API-Key header — generate keys in your platform dashboard
• Formats: PDF (with OCR fallback), DOCX, TXT, HTML, MD — up to 50 MB
• Limits: governed by your plan (Starter: 50 docs, Professional: 500, Enterprise: unlimited)
The Model Context Protocol gives Claude direct access to our legal knowledge base. Search four curated corpuses, ingest documents, and get source-cited answers — all from your editor or Claude Desktop. No API wrapper needed.
Available Corpuses
Do Better Legal
Norwegian law & family law
CO2 & Climate
EU ETS, emissions, UNFCCC
EU AI & Digital Law
AI Act, GDPR, NIS2, DSA
Telecom Regulation
EECC, spectrum, BEREC
9 MCP Tools
Use case
“Ask Claude to research EU AI Act obligations for your product. Claude searches the corpus, finds the relevant articles, and gives you a source-cited answer — directly in your editor, no browser tab needed.”
• Transport: SSH stdio (Claude Code) or SSE on port 8001 (Claude Desktop)
• Vector search: Qdrant with 768-dim nomic-embed-text embeddings
• Full RAG pipeline: ask tool combines search + LLM for cited answers
Create your own web scrapers from the platform dashboard. Point them at public data sources — government registries, regulatory bodies, public case databases. Schedule them to run daily or weekly. Your corpus stays current without manual work.
All sources are admin-approved before activation for security. SSRF-protected. Configurable depth and page limits based on your plan.
Transparent Revenue
No hidden fees. No platform taxes that grow over time. You earn 80% of every subscriber payment, paid monthly via bank transfer. A real-time dashboard shows your subscriber count, query volume, and revenue.
We take 20% to cover infrastructure — vector database hosting, embedding pipeline, AI inference, billing, and distribution. You keep the rest.
Focus on your data. We handle everything else.
Qdrant hosting with automatic indexing. No infrastructure to manage.
Documents are automatically chunked, embedded (nomic-embed-text), and indexed.
Configure scrapers that run on our servers. Keep your corpus fresh with automated updates.
Subscribers query your corpus through our UI, REST API, or embeddable chat widget.
Stripe handles subscriptions, invoicing, and payment processing. We manage everything.
Track subscriber counts, query volumes, and revenue. Full transparency on your earnings.
Every corpus on the marketplace meets these requirements.
Enough depth to provide meaningful answers.
Public domain, properly licensed, or owned by you.
Automated scraper or manual updates at least every quarter.
We review every corpus before it goes live on the marketplace.
Tell us about the corpus you want to build. We review applications within 48 hours.
Any publicly available or properly licensed documents: legislation, regulations, guidelines, standards, court decisions, research papers, manuals, and more. Content must be legally distributable.
You own your data. CorpusAI has a non-exclusive license to host and serve it to subscribers. You can remove your corpus from the marketplace at any time.
Yes. You can unpublish your corpus at any time. Existing subscribers will be notified and their subscriptions will not renew.
You earn 80% of each subscriber's monthly payment. Payouts are processed monthly via bank transfer. A transparent dashboard shows all transactions.
Corpora that haven't been updated in 6 months may be flagged as "unmaintained" and could be delisted if quality drops. Automated scrapers help prevent this.
You know how to scrape data, but you don't want to build a whole AI chat UI and vector database. Dump your data into CorpusAI, set a price, and we handle the rest.
Apply Now ↑