Features
Search multiple web engines and public communities, fetch useful pages, remove duplicates, rerank the evidence, and return the output your AI needs.
Explore the public playground before you integrate.
Multi-source discovery
Query major web engines and public conversations through one request, with one response model for your application.
Search Google, Bing, Brave, Yandex, Baidu, and DuckDuckGo through one API.
Add public conversations from X, Reddit, and Discourse-powered communities.
Use source, market, language, freshness, and domain controls for the query at hand.
Run a query and inspect the request and response before you write integration code.
Search Fusion
Fetch the content, reconcile overlapping sources, and rank the combined evidence for the request.
Turn useful result URLs into clean, readable content for downstream reasoning.
Convert source-specific results into one predictable model while preserving where each result came from.
Consolidate repeated URLs so duplicates do not waste context or appear as false corroboration.
Reorder the combined evidence around relevance, freshness, authority, and your retrieval objective.
Flexible outputs
Keep control of the evidence or let the API organize, extract, and synthesize it for you.
Receive a clean, deduplicated result list with source provenance.
Include extracted full-page content when titles and snippets are not enough.
Map themes, organize evidence, or extract the fields your application expects.
Generate an optional response from retrieved content while preserving inspectable sources.
Developer workflow
Replace separate search, community, scraping, deduplication, reranking, and AI-output integrations with one API surface.
Configure discovery, extraction depth, ranking, and output through a standard HTTP workflow.
Use the public playground to understand controls and response fields in the browser.
Follow cURL, Python, JavaScript, and Node.js examples built around the same request model.
Choose a plan for your expected retrieval volume and change it as your application grows.
Search multiple engines and public communities, fetch the evidence, and return one ranked, model-ready result set.