Space Frontiers research corpus

Access, provenance & rights

Search access and bulk data access are separate products. You can inspect records through the web app, REST API, or MCP server before discussing a licensed bulk delivery.

Search access

Results include canonical source URIs and available bibliographic metadata. Full text is returned only where it is present in the corpus.

Bulk access

Scoped by source segment, intended use, and volume. There is no anonymous whole-corpus download.

Provenance

DOI, PubMed, ISBN, patent, and source-page identifiers are retained where available so records can be traced back to their source.

Rights

Availability and permitted use vary by source and content segment. A Space Frontiers delivery does not replace third-party rights or attribution requirements.

Scholarly & reference works

The core of the corpus: 195M scholarly and reference records with metadata, abstracts, and full text where available.

Full text and abstracts

• Journal articles with abstract
  79% · 104M
• Patents with full text
  100% · 5.1M
• Encyclopedia articles with full text
  100% · 9.3M
• Books with complete text
  61% · 3.7M
• Journal articles with full text
  32% · 42M

Linked identifiers

DOI-linked works

158M

PubMed / PMC

46M

ISBN

11M

arXiv

1.4M

A 0.5% random sample of articles alone spans 16,000+ distinct journals and 3,300+ publishers, so full journal coverage is substantially wider. Publication dates reach back to the 1840s.

Community discussions

Public conversation streams complement the scholarly record with current, informal knowledge.

Text volume

Estimated token counts by segment, assuming ≈4 characters per token. The corpus totals roughly 7.3 trillion characters, or ≈1.8 trillion tokens.

Languages

50+ languages appear in language-tagged content (scholarly works and Telegram messages). Untagged Reddit content is predominantly English.

• English
  56%
• Russian
  30%
• Arabic
  1.9%
• German
  1.7%
• Persian
  1.5%
• Spanish
  1.5%
• French
  1.1%
• Ukrainian
  0.7%
• Portuguese
  0.5%
• Other (40+)
  5.1%

Publication dates

Scholarly coverage is deep — a third of works were published in the 2020s, yet the corpus reaches back past 1900. Community content concentrates in recent years: 63% was published within the last five years.

Embeddings & storage

Every document is chunked and embedded for hybrid retrieval: 3.48 billion passages carry both dense 2,560-dimension binary-quantized vectors and learned sparse vectors over a ≈106,000-term vocabulary. 82% of passages embed full document content, 13% condensed short-document views, and 5% abstracts.

Methodology

Figures come from block-level random samples and planner statistics taken against the production database on July 22, 2026. Token counts are estimated at ≈4 characters per token; journal and publisher counts are lower bounds observed in samples. All values are rounded and refreshed periodically as the corpus grows.