Frequently Asked Questions

Quotewise is a semantic search engine for verified quotations. Instead of matching keywords, it uses vector embeddings to find quotes by meaning. The platform includes 611K quotes from 32K originators with source documentation through QuoteSightings. It works on the web, through an API, and inside AI assistants like Claude via MCP.

Quotewise was founded by Christopher Q. Alexander (CQA) in October 2024. CQA is a detail-oriented engineer with a background in curating human knowledge systems. The platform grew from a personal quote collection into a mission to preserve authenticated human wisdom in the age of AI-generated content.

Quotewise currently has a growing corpus of 611K quotes from 32K originators across dozens of languages. The corpus grows through ongoing curation, community contributions, and partnerships with curated platforms. Every quote includes source documentation via QuoteSightings.

Semantic search converts your query and every quote into mathematical representations of meaning called vector embeddings (1024 dimensions). When you search for "courage during setbacks," the system finds quotes matching that concept even if they use completely different words. Think of it as matching the melody of an idea, not the exact notes. Results show real quotes from our curated corpus, not AI-generated text.

The Model Context Protocol (MCP) server lets AI assistants like Claude search Quotewise directly. When you use Claude Desktop or Claude Code, it can find quotes by meaning, verify attributions, and explore authors without leaving your conversation. The MCP server provides the same semantic search available on the website, accessible to any MCP-compatible AI assistant. See the MCP Tools Reference for available capabilities.

Quotes come from primary sources (books with ISBN and page numbers, speeches, interviews, letters, manuscripts), curated platforms (WikiQuote, Goodreads), contemporary voices (social media, podcasts), and pre-digital reference works (Bartlett's, Oxford Dictionary of Quotations). Every quote has QuoteSightings showing where we found it: URLs, books, platforms, and dates.

Every quote is classified using 8 attribution types: Direct (primary source exists), Attributed (multiple sources agree), Popularized (famous spreader, different originator), Paraphrased (meaning preserved, words shifted), Disputed (conflicting evidence), Apocryphal (probably not authentic), Misattributed (wrong person credited), and Traditional (cultural wisdom, no individual author). We're building numerical confidence scoring (0.0-1.0 scale) launching in 2026.

When we discover a misattribution, we correct the record and maintain history. Misattributed quotes stay in the database to correct the record — "Insanity is doing the same thing and expecting different results" stays listed under Einstein with a misattribution flag so people searching for it find the correction. Anyone can flag a misattribution with evidence, and we review all reports.

The Quotewise API is available at api.quotewise.io with full API documentation. Free tier provides basic access for experimentation. Paid tiers provide higher rate limits and deeper data access including full citation details. Authentication uses OAuth 2.0. See our plans page for current pricing and tier details, or visit the developer portal to get started.

Basic search on the website is free. The MCP server provides free queries for anonymous users. Paid plans provide unlimited web search, higher MCP and API quotas, and full citation details. The Quotekey identifier is always free and open. See our plans page for current pricing and tier details.

ChatGPT generates text that sounds like quotes — sometimes real, sometimes hallucinated, with no way to tell which. Quotewise finds and verifies existing quotes from a curated corpus with source documentation. Every result comes from somewhere real with QuoteSightings evidence. That's the core difference: retrieval, not generation. You see the same evidence we see.