deepxiv/deepxiv_sdk/deepxiv-fa-grants

DeepXiv · Grants (grants)

~1.31M research awards: ~637k NSF awards (one per award id), ~570k NIH RePORTER projects (multi-year funding merged per core project number, with per-year amounts) and ~107k EU CORDIS projects (FP5 to Horizon Europe, amounts in EUR). Synced daily; exact counts are in corpus (live). Organisation / country / PI / directorate / activity code / dates / amounts come from the source; topics, methods, domain and one-line summaries are model-generated (title-only where no abstract exists — see _enrich.basis).

Corpus: 1,313,241 documents, updated 2026-10.

Setup

pip install -U deepxiv-sdk      # deepxiv registers a token on first use

With a registered key (https://data.rag.ac.cn/register), search and read are fair use: 60/min, 5 at once, 5,000/day; on the auto-registered token they spend the deepxiv daily limit at the credits listed below. ask draws on the daily ask allowance (resets 00:00 UTC; a site-wide daily total can run out early at peak times), or 3 lifetime trial asks on the auto-registered token. On 429, back off for Retry-After seconds. Data goes to stdout; cost, quota and sources go to stderr. Exit 0 ok, 1 API error, 2 usage error. The general CLI guide is the deepxiv-cli skill.

Start here

deepxiv fa search grants "single-cell transcriptomics" -F agency=nih -F "fy>=2020" -k 5
deepxiv fa search grants "quantum computing" -F agency=nsf -F "amount_total>=1000000" -k 5
deepxiv fa read grants nih:R25HG007515 -l brief
deepxiv fa ask "Which NIH institutes fund single-cell atlas projects, and how much?" -d grants

Free samples: ids nih:R25HG007515; queries single cell transcriptomics.

Search

deepxiv fa search grants "query" [-F FILTER]... [-k N] [--offset N] [--mode M] [--head] [--json]

Modes: hybrid, dense, bm25. Up to 100 hits. Cost 2 (with --head: 3). An empty query is rejected; use "*" to list by filters only. Filters combine with AND; an undeclared field exits 2.

Advanced filters (see deepxiv fa spec grants): org_state, org_name, pi_names, topics, programs, core_project_num, is_renewal.

Read

deepxiv fa read grants ID [ID...] [-l brief|detail|full] [-e EXTRA] [-p k=v]... [-f json|md|html]

At -l full, -p section=NAME reads one section and -p characters=N the first N characters.

Facets

deepxiv fa facets grants FIELD lists the legal values and counts, free. Use it instead of guessing a spelling. Fields: agency, research_domain, ic_or_directorate, activity_code, org_country, org_state.

Ask

deepxiv fa ask "question" -d grants [-e low|medium|high] [-p k=v]... [--no-stream] [--json]

Agentic Q&A over this domain: the model turns the question into filters plus keywords from the domain's own self-description, reads the hits and answers with [grants:<grant_id> <title>] citations; /stream is NDJSON with real token deltas

Always pass -d grants: automatic domain picking is still early.

What this corpus will and will not tell you

• One NIH document = all yearly awards under a core project number (amount_total is the sum, funding_by_year the breakdown); one NSF document = one award; one CORDIS document = one project (grant_id = cordis:<project id>), org_* is the coordinator and amount_total the EU contribution in EUR (currency=EUR at level=head; NIH / NSF have no currency field and are in USD).
• org_country casing is inconsistent in the source (both UNITED STATES and United States, 1.088M awards) — pass both spellings.
• topics are open labels with tens of thousands of values: good to read, poor to filter on — use research_domain.
• Where no abstract exists the summary comes from the title alone (_enrich.basis=title_only).
• The index holds no full text: level=full is not available yet (abstracts are already in level=detail; progress reports and publications are not in this corpus).

Citing

Cite records as [NIH R25HG007515]. Say plainly when the corpus has nothing.

Generated by scripts/gen_fa_skills.py from deepxiv fa spec grants --json on 2026-10-19. Field names and prices can change; deepxiv fa spec grants is the live source.