Truey 作者： CodKid1710

Truey adds a research summary card to your search results when your question
looks scientific: "does coffee increase blood pressure," "does exercise
improve memory," and so on. Ask a question on Google, Bing, DuckDuckGo,
Brave, Ecosia, Startpage, or Yahoo, and Truey:

1. Fetches relevant peer-reviewed papers from PubMed, Semantic Scholar,
   Europe PMC, arXiv, and/or OpenAlex
2. Filters out anything that doesn't actually match your question. Truey
   would rather show "limited data" than an irrelevant paper
3. Summarizes the evidence using an AI model of your choice, highlighting
   the single clearest finding and rating overall confidence as Low,
   Moderate, or High

KEY FEATURES

• Inline search cards: no extra clicks, the summary appears alongside
your normal results
• Multi-source retrieval with relevance ranking, so off-topic papers never
reach the summary
• Bring your own AI provider: Groq, OpenAI, Anthropic (Claude), Grok, or
run it fully locally with Ollama / llama.cpp (no API key, nothing leaves
your device)
• Adjustable reading level: plain English up to expert/researcher level,
globally or per scientific field
• Evidence confidence rating on every summary
• Privacy controls: local-only mode, and you can disable history/caching
entirely

YOUR DATA

Truey has no backend server of its own. Requests go directly from your
browser to the paper databases and AI provider you choose. Settings,
history, and cached abstracts are stored locally in your browser only,
never sent anywhere except the provider you configure.

This is an independent, open-source project, not affiliated with any of
the listed AI providers or research databases.