Danubia
Show HN: EU-hosted content extraction API
Last verified:
What is Danubia?
Danubia is a clean text extraction API that converts URLs into well-formatted markdown content, removing boilerplate like navigation, ads, and cookie banners. It's designed for feeding LLMs with fresh web data (RAG, agents, evaluation sets), building products on scraped content, and automating large-scale data extraction from the web.
Danubia pricing
Pricing model: Freemium
Free during beta with 500 credits; credit-based billing model coming when launched
Danubia pros
- Best-in-class content extraction benchmarked against Trafilatura, Readability, Turndown, and Firecrawl
- Polite, production-grade crawler respecting robots.txt with per-domain rate limiting and deduplication across users
- Optimizes token usage for AI pipelines with cleaner markdown input resulting in better LLM answers
- Batch API for large-scale scraping (JSONL input/output) and free tier with 500 credits during beta
Danubia cons
- Currently in beta testing phase; not yet production-ready
- Only extracts clean text/markdown; structured data extraction is a planned future feature
- Requires users to provide URLs; built-in crawling/discovery not yet available
Frequently asked questions about Danubia
How does Danubia compare to other extraction tools?
Uses a custom-built extraction library rebuilt from scratch for speed and accuracy, with benchmarking showing performance at least equal to or better than Trafilatura, Readability, Turndown, and Firecrawl on most pages.
How does Danubia handle rate limiting?
Operates like a large-scale search engine crawler with per-domain concurrency limits, crawl delays, and request deduplication—if multiple users request the same page within a short window, it's only fetched once.
Is there a batch API?
Yes, submit a JSONL file of URLs and receive clean text back in JSONL format for large-scale scraping jobs.
What features are planned?
Structured data extraction API accepting JSON schemas to extract and shape data exactly as requested, plus built-in crawling to discover and map website content automatically.