Research
Search the live web, compare multiple domains, and surface the passages most relevant to your question.
Private web retrieval
NoemaSearch
The private search infrastructure behind Noema. Search directly here—or use the app to research, open, and verify live sources with citations.
Explore research in NoemaIn the Noema app · macOS and iOS
Web research that shows its work.
Noema searches the live web, reads the sources it finds, identifies the most relevant evidence, and provides citations you can open and verify.
12 research capabilities, organized around three core actions.
Open
Read beyond a search snippet in articles, public text, documentation, and text-based PDFs.
Find
Return to a source and locate a specific phrase, claim, method, result, or limitation inside it.
12 built-in capabilities
From question to verifiable evidence.
Search finds the source. Noema reads it, locates the evidence, and keeps the path back to the original material intact.
Research
Discover current sources and decide which evidence matters.
- 01
Live web research
Search for current information across multiple web sources.
- 02
Multi-source comparison
Gather evidence from several domains instead of relying on one result.
- 03
Relevant passage ranking
Prioritize passages using the query, titles, headings, and phrase matches.
- 04
Search controls
Use SafeSearch and optional day, week, month, or year freshness filters.
Open
Read the source itself, not only the text shown in search results.
- 05
Article reading
Extract the readable content from ordinary HTML articles.
- 06
Plain-text reading
Understand public text documents and text-based pages.
- 07
PDF research
Read PDFs with selectable text and preserve page-number evidence.
- 08
Source opening
Open a returned source to read beyond the initial evidence.
Find
Trace an answer back to its exact, inspectable location.
- 09
Evidence finding
Locate a word, phrase, or claim inside a previously returned source.
- 10
Precise evidence locations
Show headings, HTML line ranges, or PDF page numbers.
- 11
Verifiable citations
Turn citation markers into deterministic, tappable source links.
- 12
Evidence transparency
Label sources as read, snippet-only, blocked, unsupported, timed out, or unavailable.
Primary-source research
Research papers, not just web pages.
Give Noema a topic, paper title, arXiv link, or public PDF URL. It can move from discovery to page-level evidence without reducing the paper to a search-result snippet.
Example
“What evidence does this arXiv paper provide for its main conclusion?”
Noema can read the methodology and results, identify the most relevant passages, and answer with a citation such as [1, p. 7].
“Find where the authors discuss limitations.”
Noema searches within the same paper instead of starting over.
From link to page-level evidence
- 01
Find relevant papers through web search.
- 02
Read an arXiv abstract page.
- 03
Download and extract text from a public paper PDF.
- 04
Preserve page numbers while reading.
- 05
Rank the passages most relevant to your question.
- 06
Find a specific term, method, result, or limitation.
- 07
Open additional sections when more context is needed.
- 08
Cite the paper with a tappable link and page-level evidence.
Public source coverage
Built for the open research web.
The same workflow works across publicly accessible source types—where the page or file can be read without authentication.
- 01
arXiv papers
- 02
PubMed abstracts and PubMed Central articles
- 03
University research repositories
- 04
Government and regulatory reports
- 05
Technical documentation and white papers
- 06
Public reports and HTML-described datasets
- 07
Public standards and specifications
- 08
News articles and long-form analysis
Private by design
Research without leaving a profile behind.
NoemaSearch is a dedicated SearxNG instance maintained for the Noema community. It combines open metasearch indexes while separating your browser and device from the services that provide results.
The same private endpoint powers live web retrieval in Noema and is available directly in any browser, free to use and without authentication.
No identity layer
NoemaSearch does not require an account, issue a tracking token, or attach a persistent identifier to a query.
No retained research trail
Search queries, visited URLs, and extracted page text are not persisted on NoemaSearch infrastructure.
Short-lived processing
Documents are processed temporarily and cached only in memory for a short period.
Search directly. Research deeply.
No account for search. Available in Noema on macOS and iOS.