Summary
The distinction here is licensing: the content behind the answers is rights-cleared and attributed to the publisher it came from.
Two tools cover both halves of the job — a live query answered from the web and market index, and a recommendation call that returns articles with source, date and relevance attached. Which body of content you get is decided by the model id you pass, so the same tool serves general web search, market data or one publisher's archive.
What it is
Dappier's own server over its rights-cleared data models: one tool answers a live query, the other returns recommended articles from a chosen content model.
What you get
- Direct answers to real-time questions — breaking news, weather, travel, deals, trending topics — from an AI-powered web index
- Stock market coverage in the same call shape: prices, financial headlines, trade signals and market summaries built on live Polygon.io data
- Content recommendations from domain-specific models, spanning sports, lifestyle and publisher-specific sources, selected by the model id you pass
- Structured results rather than prose: title, summary, publication date, source and domain, image preview where there is one, and a relevance score
- Control over how recommendations are picked — search algorithm choices of most_recent, semantic, most_recent_semantic or trending, a reference domain to favour, and how many results come back
- A marketplace of further data models, addressed by id through the same two tools
Requirements
A Dappier API key from platform.dappier.com, passed as DAPPIER_API_KEY. Published on PyPI as dappier-mcp and launched with uvx over stdio.
Setup effort
One command plus a key — pip install dappier-mcp, then supply credentials