Ragy.ai is a web search engine and summarization api. We retrieve the search results for a query (with 10 different search engines), while we spider the top web pages and summarize them at the same time.
This makes it possible for LLM's and AI applications to retrieve up to date, really specific and accurate results.
Give it a try for free (3.000 requests per month).
RAG (Retrieval-augmented generation) enhances the performance of large language models (LLMs) by integrating them with external knowledge sources. In case of Web Rag with web search.
For each api query, we retrieve 10 sources with up to 50 web results each. We deduplicate, re-rank and clean those results while we retrieve and summarize the relevant web pages at the same time.
We use our own search engine, content specific sources like wiki and geographical search engines and generic web search engines like Google, Bing and Brave, so in total over 10 sources for each search
For all results we provide an url, title and description (like most search engines) and a content summary for the most important pages.
We summarize each html page into the most important unique sentences (with the textrank algorithm).
Because everything is done simultaneously, an api call normally takes max 1.5 seconds (you can set a max from 0.5 to 5 seconds).
Start for free with 3.000 requests a month. Need more? The price is € 5,- per 1.000 requests. Go to RapidApi for pay per api call. Please contact us with any specific needs. Working on an open source application or open access publication? Contact us to apply for a free subscription.