# NobleBlocks — full guide for language models

> Open access academic publishing and social networking for researchers. Search 300M+ papers, run AI literature reviews, peer review, and publish — all in one platform.

NobleBlocks is a free research platform: search across 300M+ scholarly papers, run a literature review, and read each paper's abstract, authors, identifiers, citations and related work. Membership is free and there are no paid plans at the moment.

## What is on the site

- Paper search over the NobleBlocks paper corpus (plain search results never wait on an AI step).
- Literature review: PICO suggestion, screening, data extraction, risk of bias, evidence synthesis; Deep Review runs it iteratively.
- Topic, field, institution and author pages built from the corpus.
- Member profiles and works published on NobleBlocks.
- An MCP server for AI assistants (see "MCP" below).

## Key URL patterns

- Paper: https://www.nobleblocks.com/publications/<id>/<slug> — <id> is a stable id: an OpenAlex work id (W…), "pdb-<n>" for other corpus rows, a trial-registry id (NCT…) for registrations, or a UUID for works published on NobleBlocks. A URL that uses a DOI or a NobleID redirects permanently to this form when the paper has a stable id.
- Member profile: https://www.nobleblocks.com/@<username> — name, bio, affiliation, ORCID and the member's public works.
- Author (from the paper corpus): https://www.nobleblocks.com/researcher/<ORCID iD>
- Topic: https://www.nobleblocks.com/topics/<topic-slug>
- Field: https://www.nobleblocks.com/fields/<field-slug>
- Institution: https://www.nobleblocks.com/institutions/<institution-slug>
- Blog post: https://www.nobleblocks.com/blog/<slug>

## Identifiers

- DOI: shown on paper pages and linked to https://doi.org/<doi>.
- NobleID: the persistent identifier issued by NobleID (https://www.nobleid.org). A paper page shows "NobleID" with the code (e.g. NI…) only when the work is minted and NobleID confirms the record matches the paper; the code links to the work's ARK resolver URL (https://www.nobleid.org/ark:/<NAAN>/<path>). The ARK is also in the page's JSON-LD (identifier, sameAs) and in the dc.identifier meta tag.
- "NobleBlocks ID" (NB…): a NobleBlocks-internal id for works uploaded on NobleBlocks. It is not a NobleID and does not resolve on nobleid.org.

## What each page type contains

### Paper page — https://www.nobleblocks.com/publications/<id>/<slug>
- Server-rendered: title, abstract, authors (with ORCID where known), venue, publication date (a year only when that is all the source gives), DOI, NobleID when confirmed, citation and reference counts, open-access PDF link when one exists.
- <link rel="canonical"> is the stable-id URL; Google Scholar citation_* meta tags; JSON-LD ScholarlyArticle (headline, author, datePublished, identifier with DOI and ARK/NobleID, sameAs https://doi.org/… and the ARK URL, isPartOf Periodical) plus BreadcrumbList.
- Clinical-trial registrations are marked as registrations (JSON-LD MedicalTrial), not as papers.

### Member profile — https://www.nobleblocks.com/@<username>
- Server-rendered: display name, bio, affiliation and a list of the member's public works with links.
- JSON-LD ProfilePage whose mainEntity is a Person (name, alternateName @username, affiliation, identifier ORCID, sameAs ORCID / the person's own links / NobleID person record when it resolves), hasPart lists the works; BreadcrumbList.

### Author page — https://www.nobleblocks.com/researcher/<ORCID iD>
- Server-rendered: name, affiliation, paper and citation counts in the NobleBlocks index, the author's most-cited papers.
- JSON-LD ProfilePage → Person with ORCID identifier and sameAs; BreadcrumbList.

### Topic / field / institution pages
- https://www.nobleblocks.com/topics/<slug>, https://www.nobleblocks.com/fields/<slug>, https://www.nobleblocks.com/institutions/<slug>: server-rendered descriptions, counts and most-cited papers.
- JSON-LD CollectionPage (with the institution as an Organization where relevant) and BreadcrumbList.

### Blog — https://www.nobleblocks.com/blog
- JSON-LD Blog / BlogPosting and BreadcrumbList.

- [NobleBlocks MCP Server: Search 300 Million Academic Papers Directly from Claude](https://www.nobleblocks.com/blog/nobleblocks-mcp-server-search-papers-from-claude): NobleBlocks is now available as an MCP server for Claude Desktop and any MCP-compatible AI assistant. Search, filter, and discover research papers without leaving your AI workflow.
- [We built an AI research search that doesn't make up papers](https://www.nobleblocks.com/blog/ai-research-search-without-hallucinations): Most AI tools confidently cite papers that don't exist. We took a different approach — every result links to a real, verifiable paper. Here's how.
- [Stop writing Boolean queries — just ask what you want to know](https://www.nobleblocks.com/blog/natural-language-search-for-academic-papers): You shouldn't need to be a search expert to find papers. Type your question in plain English and get back studies that actually answer it.
- [A Letter to Our Community](https://www.nobleblocks.com/blog/a-letter-to-our-community): An honest update from the founder: how we kept NobleBlocks alive through the build, exactly where the liquidity went, and the public, verifiable commitment to pay every cent back.
- [Systematic reviews in days, not months](https://www.nobleblocks.com/blog/ai-systematic-literature-review): The screening and extraction phases of an SLR are mechanical torture. Our AI assistant handles the spreadsheet work so you can focus on the science.
- [Deep Review: let an agent do your literature search](https://www.nobleblocks.com/blog/deep-review-agentic-research-workflow): For questions that need more than one search query, Deep Review builds a research strategy, runs it, and hands you a fully-cited briefing.
- [We index 300 million papers. Here's what that covers.](https://www.nobleblocks.com/blog/300m-papers-and-counting): Every major source — OpenAlex, PubMed, arXiv, bioRxiv, CrossRef — in one search. No discipline left out. Refreshed daily.
- [Follow the conversation: citation graphs for paper discovery](https://www.nobleblocks.com/blog/citation-graphs-explained): Every paper cites earlier work and gets cited by later work. Following those threads is the fastest way to find related research you'd otherwise miss.
- [Ask a paper your question instead of reading all 12 pages](https://www.nobleblocks.com/blog/ask-follow-up-questions-on-any-paper): You usually need one specific thing from a paper — sample size, dose, statistical test. Just ask. Get a cited answer in seconds.
- [Chat with any PDF instead of reading it front-to-back](https://www.nobleblocks.com/blog/pdf-chat-read-smarter): Drop in a paper — from our index or your own files — and have a conversation with it. Get answers with page numbers, not generic summaries.
- [We always show you the free version of a paper (when it exists)](https://www.nobleblocks.com/blog/open-access-first): Paywalls shouldn't stop you from reading publicly-funded research. We find and surface every legal free copy — PubMed Central, preprints, institutional archives.
- [How we compare to Elicit, SciSpace, Consensus, Scite, and Semantic Scholar](https://www.nobleblocks.com/blog/nobleblocks-vs-elicit-consensus-scite-semantic-scholar): An honest look at what each AI research tool does well, where they fall short, and what we do differently.

## How to cite a page

- Cite the paper itself, with its DOI (https://doi.org/<doi>) and, when the page shows one, its NobleID ARK URL.
- When citing NobleBlocks as the source, use the page's canonical URL (its <link rel="canonical">) and the name "NobleBlocks".

## Crawling

- Allowed: content pages listed above, https://www.nobleblocks.com/llms.txt, this file, the sitemaps.
- Not allowed (robots.txt): https://www.nobleblocks.com/api/, search result URLs (https://www.nobleblocks.com/search?q=…), account pages (settings, library, history, inbox, billing) and the interactive AI tools (research assistant, Deep Review chat, notebooks). These would start paid model calls or hold nothing to index.
- Sitemaps: https://www.nobleblocks.com/sitemap.xml and https://www.nobleblocks.com/publications-sitemap.xml.
- Languages: localized pages accept `?lang=<code>` (en, es, id, fr, de, zh, ja, ar, ko, tr, sv, ru, he, uk, nl, pt, it, hi, pl, vi, th); each declares hreflang alternates and its own canonical.

## MCP

- `pip install nobleblocks-mcp` — tools: search_papers, get_paper, find_similar, get_citation_graph, search_by_entity, create_literature_review. Docs: https://www.nobleblocks.com/docs/mcp

## Contact

- https://www.nobleblocks.com/contact · https://www.nobleblocks.com/support