AI search for evidence-backed research and learning
Metaphor is an AI-powered research search engine that returns source‑attributed answers and summaries for web pages and video transcripts. It suits researchers, students, and knowledge workers who need quick, citation-ready explanations rather than raw links. Metaphor offers a usable free tier with query limits and paid plans for heavier use, making it a practical choice for individuals and small teams seeking research-centric search rather than a general chatbot.
Metaphor is an AI research search engine that finds, summarizes, and cites web content to support fast learning and fact-checking in the Research & Learning category. Its primary capability is retrieval-augmented search that surfaces short, source‑attributed answers with links to the original content. The key differentiator is its focus on evidence-first responses (answer + provenance) rather than open-ended chat. Metaphor serves students, product researchers, and journalists who need quick, citeable summaries of web pages and videos. Pricing is accessible: a free tier exists with daily query limits and paid Pro/Team plans to unlock higher monthly quotas (pricing noted below, approximate).
Metaphor is positioned as an AI-first search engine focused on research and learning rather than conversational general-purpose AI. Launched by a small team building retrieval-augmented models and a web-scale index, Metaphor aims to reduce time spent sifting through links by returning concise, evidence-backed answers with direct links to sources. The product emphasizes provenance: every answer includes inline citations and direct links so users can verify claims quickly. It markets itself to knowledge workers who prioritize verifiable summaries over freeform chat, distinguishing it from chatbot-centric tools in the research-learning space.
Metaphor’s core feature is source‑attributed search: enter a query and it returns a short answer paragraph followed by ranked source snippets and links. You can also paste any URL to get a one-click summary of that page’s contents and main points. The tool supports summarizing YouTube videos via URL by extracting and condensing transcripts into bullets and time-stamped references. A browser extension (Chrome) brings same-page quick summaries and inline search, while saved queries and basic workspace folders let users store and revisit research. Metaphor additionally exposes an API (beta/paid) so teams can integrate its retrieval and summary endpoints into apps or pipelines.
Pricing splits between a free tier, an individual paid plan labelled Pro, and enterprise/Team options. The free tier (approx.) provides limited daily searches and URL summaries intended for casual or exploratory users. Pro is a monthly subscription (approx. $10–$15/month) that increases monthly query allowances, removes some rate caps, and unlocks API keys for personal use. Team/Enterprise is quoted as custom pricing and typically adds shared workspaces, SSO, higher API quotas, and contract terms. Exact quotas, rate limits, and API pricing change frequently; check Metaphor’s pricing page for the current numbers (some price figures above are approximate).
Real-world users include researchers and practitioners who need fast, verifiable answers. Example use-cases: a product manager using Metaphor to summarize competitor blog posts and extract 5 concise competitive insights; an investigative journalist using it to gather source‑linked summaries from dozens of public reports to speed fact-checking. Academic students use Metaphor to convert assigned readings and lecture videos into concise notes. Compared with Perplexity AI, Metaphor leans more heavily on a search-first UI and direct source surfacing rather than a chat-oriented conversational interface.
Three capabilities that set Metaphor apart from its nearest competitors.
Current tiers and what you get at each price point. Verified against the vendor's pricing page.
| Plan | Price | What you get | Best for |
|---|---|---|---|
| Free | Free | Limited daily/weekly searches and URL summaries, basic browser extension | Casual learners and trial users |
| Pro | $12/month (approx.) | Higher monthly search quota, personal API key, faster rate limits | Individual researchers and students |
| Team | Custom | Shared workspaces, SSO, larger API quotas, enterprise SLAs | Small teams and organizations |
Copy these into Metaphor as-is. Each targets a different high-value workflow.
Role: You are an evidence-first research assistant using Metaphor to extract and cite facts. Task: Summarize the specified web page URL into a 150-word evidence-first summary that directly cites the original page (include URL). Constraints: 1) Write one paragraph (max 150 words). 2) Include 2 short bullet takeaways after the paragraph. 3) Add the full source citation (title, author if available, date, URL). Output format: JSON with keys: "summary", "takeaways" (array), "citation". Example output: {"summary":"...","takeaways":["...","..."],"citation":"..."}. Now summarize this URL: [PASTE URL HERE].
Role: You are a concise product analyst using Metaphor to pull evidence-backed competitor insights. Task: Given a competitor domain or product page URL, return five distinct product insights pulled from recent public sources. Constraints: 1) Each insight is 20–30 words. 2) Attach one inline source citation (title + URL) per insight. 3) Only use sources published within the past 18 months. Output format: JSON array of objects: [{"insight":"...","source":"title — URL"}, ...]. Example insight: {"insight":"Launched free developer tier to increase adoption","source":"Blog: New Pricing — https://..."}. Provide only the JSON.
Role: You are an evidence-first fact-checker using Metaphor to cross-verify claims. Task: Given up to five URLs or search queries, produce a structured brief that lists each major claim, a 1-sentence verification verdict (True/False/Unclear), three supporting or contradicting source citations, and a 15-word rationale. Constraints: 1) Limit to 8 claims total. 2) Prefer primary sources and peer-reviewed items. 3) Note publication dates beside each citation. Output format: JSON: {"claims":[{"claim":"...","verdict":"...","rationale":"...","citations":[{"title":"...","url":"...","date":"YYYY-MM-DD"}]}]}.
Role: You are a product strategist using Metaphor to build a competitors matrix. Task: For each competitor domain or product name provided (up to 6), fetch evidence and fill a matrix with: positioning statement (20–30 words), top 3 features, recent pricing changes (last 2 years), and one supporting source per cell. Constraints: 1) Output must be a JSON array of competitor objects. 2) Use only public company pages, product blogs, or reputable press coverage. Output format example: [{"competitor":"Name","positioning":"...","features":["f1","f2","f3"],"pricing_changes":"...","sources":[{"title":"...","url":"..."}]}].
Role: You are a senior researcher using Metaphor to synthesize web and academic sources into themes and gaps. Step 1: Search provided queries or URLs and collect the top 12 relevant sources (mix web articles, reports, and preprints). Step 2: Generate 5 high-level themes (one sentence each) supported by 2–3 cited sources per theme. Step 3: List 4 concrete research gaps or unanswered questions with brief methodological suggestions. Constraints: 1) Provide publication year with each citation. 2) Limit total output to 500–650 words. Output format: JSON with keys: "themes", "gaps", "sources".
Role: You are a domain expert preparing a grant background section using Metaphor. Task: Using the input topic, locate and synthesize up to 10 high-quality public sources (peer-reviewed, preprints, government reports) and draft a 700–900 word background section that: 1) Frames the problem, 2) Summarizes key findings with inline citations (author, year, URL), 3) Identifies 3 specific knowledge gaps motivating the proposed study. Constraints: Use APA-style in-text citations and include a separate references list at the end with full citations and URLs. Example inline citation: (Smith et al., 2021).
Choose Metaphor over Perplexity if you prioritize direct source snippets and URL/YouTube summarization in a search-first interface.
Head-to-head comparisons between Metaphor and top alternatives: