How this works

Funding Radar sweeps public funding sources every day and keeps a single, searchable list of what is open. You can also type your institution's name and get calls ranked against what your institution actually publishes on.

The part that matters most is the part most tools skip: a call that matches your keywords most strongly may explicitly exclude your field. Funding Radar reads the scope text, finds those clauses, and drops the call with the sentence quoted — instead of ranking it first because the words matched.

Where it can do that, and where it cannot. The clause has to be somewhere machine-readable. It is for the EU portal, for US federal calls through Grants.gov, and for ICGEB, IDRC, NRF Kenya and ANR — 832 of the 1325 calls listed here. For the rest it is not: NWO's call page is a stub ending "please visit the website of Regieorgaan SIA", four in ten UKRI pages point at a separate portal, and SNSF, Novo and GrantConnect publish the scope as a PDF or not at all. Eighteen of those pages were fetched and read on 15 September 2026 and between them stated two exclusions, both about what the money may be spent on rather than what may be proposed. The "la Caixa" Foundation publishes its calls in full but reserves every right to them, so their text is not read here either. So on those 493 calls this tool lists the call and links you to it, and the reading is still yours to do.

All US federal agencies — NIH, NSF, CDC, FDA, USDA, DOD, NOAA

Since 1 October 2025 this is also the only official channel for NIH funding opportunities (NOT-OD-25-143); the NIH Guide now carries policy notices only.

Horizon Europe, Erasmus+, EU4Health, CBE JU, EDCTP3 and the other EU programmes

Scope text and eligibility conditions are fetched per topic, which is how exclusion clauses are detected.

Australia's whole-of-government grants registry, including the Medical Research Future Fund

Queued for review: roughly one entry in seven is research funding. Its detail pages serve their content only to a request carrying a session cookie, which any listing page sets; browsing needs no account, as its own help page states.

UK Research and Innovation councils — MRC, BBSRC, EPSRC, NERC, ESRC, Innovate UK

The feed carries only the most recent items and no deadline field.

Dutch Research Council calls across all domains

Crawled rather than fetched: NWO publishes no feed and no API. Every call is read by a person before it appears here.

Canadian Institutes of Health Research — grants, awards, fellowships and training

The only national council of ten checked whose RSS carries its own closing dates rather than leaving them on a detail page.

Life science, health and sustainability grants

Research grants and fellowships in life sciences and biotechnology

ICGEB member states are largely developing countries, so applicants are the intended beneficiaries rather than third-country partners.

Germany's main research funder — Priority Programmes, international joint calls and research training groups

Fetched by relay. DFG's servers answer requests from data centres with a bot-verification challenge and ordinary connections normally, while their own rules for automated readers permit both the feed and the call pages, and they publish an index of their pages for such readers — so the bytes are fetched on a home line and handed to the server, identifying itself the same way, one pass a day. Queued for review: the channel carries announcements as well as calls, and each call states two mandatory dates — elan registration and the proposal — of which the proposal date is the one shown.

National Research Fund calls, including its international collaborative networks

The first national council in the catalogue. Applicants are Kenyan institutions, though several calls are collaborative networks that name partner countries.

SNSF calls across all disciplines

Most SNSF schemes require a Swiss-based applicant.

Finland's national research council — Academy Projects, Research Fellowships, Centres of Excellence and the international partnerships it co-funds

Crawled rather than fetched: RCF publishes no feed and no API. It states the closing date on the listing itself, in English, which is rare — most councils leave it on the call's own page. The national portal research.fi carries only one of its nine open calls, so this is not a duplicate of anything.

France's national research agency — generic and thematic calls, plus the European partnerships it co-funds

The feed mixes news with calls, so only call pages are kept. Each call states its timeline as labelled dates, and the opening date sits in the same block as the deadline — it is reported for review rather than used.

Research funding for the Global South

Deadlines are stated in prose rather than a feed field, so many IDRC records carry no machine-readable deadline.

the French-speaking community of Belgium — fellowships, research projects and the bilateral and WEAVE partnerships

Crawled and queued for review. Its Clôture field is the date FNRS itself closes, which on a partnership call is not the date that decides whether you are in it. All four transnational calls checked on 15 September 2026 stated a Clôture seven or eight days after the mandatory pre-proposal deadline at the call secretariat — RAMP 29 vs 22 September, Biodiversa+ and Water4All 18 vs 10 November, DUT 24 vs 17 November — and all four are published here with the secretariat's date. The bilateral calls were checked too and are correct as stated: there each side submits to its own agency.

Ireland's national research funder — the former Science Foundation Ireland, Irish Research Council and Enterprise Ireland research programmes, plus the European partnerships it co-funds

Crawled rather than fetched: Research Ireland publishes no feed and no API, and its own /api/ path is disallowed. Queued for review because the deadline field names its own kind and not every kind is a proposal deadline — Global Talent Ireland states an Expression of Interest date in it. The kind is shown beside the date so a person can tell. Eleven of the seventeen live calls carry no readable date: four are marked Rolling and six have not announced one yet.

India's medical research council — investigator-initiated extramural grants, expressions of interest and centre-of-excellence calls

One of four sources in this catalogue whose own text may be republished, and the only one outside the EU, the United States and the United Kingdom. Each row links to the call document itself because ICMR publishes no per-call page. Its results announcements share the listing with its calls and are filtered out. Every call is read by a person before it appears here.

Requests for applications in science and technology

Luxembourg's national research funder — 27 funding instruments, from PhD grants to industrial fellowships

Read from the WordPress post-type feed: the REST route for the type is closed and the page renders its deadline in JavaScript. Queued for review rather than published straight, because instrument pages quote partner agencies' closing dates alongside their own.

Poland's national funder of basic research — OPUS, SONATA, SONATA BIS, MAESTRO, DIOSCURI, and the Weave, bilateral and Biodiversa+ partnerships it co-funds

NCN states each closing date as a machine-readable date in its own listing, and that is the date shown. The link goes to the Polish announcement, which is the only one NCN links to; eight of the eleven calls listed on 15 September 2026 also had an English version at the same path under /en.

Spain's national research agency — the Plan Estatal calls: Generación de Conocimiento, Ramón y Cajal, Juan de la Cierva, FPU, Severo Ochoa / María de Maeztu and the PCI international collaboration rounds

Crawled rather than fetched: AEI publishes no feed and no API. Its padlock badge is not read, because 'Cerrada' there means 'not open today' and includes calls that have been announced but have not opened yet — on 15 September 2026 all ten 2026 calls read Cerrada while Ramón y Cajal was live with a 10 December deadline. The closing date decides instead. Deadlines are read from the machine-readable <time> attributes of the application window, taking the closing end, because the opening date sits under the same label and a pre-announced call has both in the future.

Brazil's national council — chamadas across all fields

Crawled and queued for review. The submission window is stated in one labelled line of the page body, in Portuguese.

Portugal's national research funder — PhD studentships, scientific employment, R&D projects, advanced computing, and the bilateral and European partnerships it runs

Crawled and queued for review. FCT states its closing date on every call page with the date before its label: "30 setembro 2026 (17h) / Encerramento das candidaturas". Each line of the call's timeline is read label first, so that is the date shown. The dates on the listing itself are not used: it draws the day as an image and writes only the month in text, which would round every deadline up to the end of its month.

Türkiye's national research council — its open calls, which on the day of writing are all industrial R&D programmes (1501, 1507, 1707, 1711, 1812/BiGG, 1831, 1832) rather than the academic ARDEB schemes

Read from the Turkish page: the English mirror is the same view against the translations and carries one of the seven calls. The application window is stated as a range with a single label in front of it, so the reviewer sees both the opening and the closing date in the evidence and picks.

Mauritius's national research and innovation council — its grant schemes, from proof of concept to social innovation

Status only, never a date. MRIC states a closing date on one page, /apply, and only while a round is advertised; the schemes index that is crawled says per scheme whether the Council is accepting applications and never gives a date. Expect this source to show nothing open for most of the year — that is the site being accurate, not the crawler failing.

The research and innovation calls listed on caixaresearch.org, the "la Caixa" Foundation's biomedical and health research site: health research projects, health innovation (CaixaImpulse) and healthcare research

A major private funder of biomedical research in Spain and Portugal, read from its own list of calls. The closing date shown is the one the foundation states, and a call whose closing date has passed is not listed as open. Calls are led by public or non-profit organisations in Spain or Portugal. A person checks each call, and each new round of it, before it appears here.

Chile's national research agency — Fondecyt, Becas Chile, Fondef and Fondap, telescope time on the Chilean share of the international observatories, and its bilateral programmes

Queued for review, and only the six most recently opened rounds are visible: the listing turns its remaining pages through a route its own rules put out of bounds to automated readers, so the archive is out of reach. Each card states an opening date and an expected decision date beside the deadline; only the field ANID labels Cierre is read.

Spain's national health research funder ISCIII, including the Acción Estratégica en Salud; AGAUR, the Catalan agency that funds the university and research system of Catalonia; the CNIC, the national cardiovascular research centre, whose Cicerone placements and master's scholarships are aimed at students; and the CSIC, Spain's largest public research body. All of them read through Spain's national subsidy database, where every public subsidy call — state, autonomous-community and local — must be published by law

Read by organ, never by keyword: BDNS holds 656,110 records and a search for "investigacion" returns 6,715 of which a sample of 22 held no competitive call at all, being municipal science prizes, university internal resolutions and direct awards. The database's own `abierto` flag is not read either: all fifty most recent ISCIII and AEI records report it false, including calls closing weeks later, so the application window decides. AGAUR is read here because its own site permits no automated reading at all, while Spanish law requires it to publish the same calls in this database. ISCIII publishes no usable call listing of its own: the page at /convocatorias is staff recruitment. The AEI is left to its own source because this database states its calls as "Resolución de 25 de septiembre de 2026 de la Presidencia..." where its own pages say "Ayudas para contratos Ramón y Cajal (RYC) 2026", and reading both would list every one of them twice. Three further research bodies are polled nightly and will be listed here only once they publish: CNIO, whose most recent call in this database dates from 2018, CIEMAT, whose most recent dates from 2023, and INIA, which has none at all.

South Africa's medical research council — the National Health Research Enterprise calls, early-investigator awards and PhD scholarships

Crawled rather than fetched: SAMRC publishes no feed and no API. The listing states each deadline as "Friday, October 9, 2026", month first, which this project's parser refuses on purpose; the date shown here is read from the call's own page, where the same field is rendered day-first. Every call is read by a person before it appears here.

Norway's national research funder — researcher projects, research centres, innovation projects for industry, and the transnational partnerships it co-funds

Not collected, and the zero beside this source says so rather than saying Norway has nothing open. The site refuses this register's server as a robot, on pages the site's own rules for automated readers allow. Access for one address was requested through the Research Council’s answer service and declined on 29 October 2026. The listing could be read by presenting the crawler as a browser, which this register does not do, here or anywhere: a register that reads sites which have refused it is not one to trust about eligibility. So Norway is absent by the funder’s decision and this register’s, and the recipe is kept against the day either changes.

Some obvious-looking sources are missing on purpose. Listing them here so the omissions read as decisions rather than gaps.

• Not authoritative. Matching flags what to confirm, and it never confirms that you are eligible. It rules a call out on eligibility in one case only: the call's own text says an applicant from outside the funder's country may not apply, and the register knows both countries. Every other clause about who may apply is quoted for you to read, because whether it bites depends on things the register is not told. Always read the call text before committing time to an application.
• Not complete. Several relevant funders publish no machine-readable feed at all. Absence from this list does not mean absence of funding.
• Not a substitute for the funder's own page. Deadlines change, calls get amended, and some sources state deadlines only in prose. Where a deadline was inferred rather than published as data, the record says so.

The author has written a book on what happens to research funding selection when a credible application becomes cheap to write: Grant Funding in the Age of AI (2026, ISBN 979-8-163978-82-3). One of its findings bears on this tool directly. Ranking a shortlist is worth less, against simply drawing from it, than most people assume, because two equally careful readings of the same proposals agree less than most people assume. The ranking here is a starting point for your own reading rather than a verdict. The figures and the scripts that draw them are open at doi.org/10.5281/zenodo.23478426.

There is a paper describing this register: what it holds, where each record comes from, how the closing dates are read, and what is known to be wrong with them. It is deposited with a DOI, so it can go into a proposal, a report or a reference list.

Kahlem, P. (2026). Funding Radar: an open register of public research funding calls, with machine-read closing dates and per-record reuse terms (snapshot of 5 October 2026) [Data set and resource description]. Zenodo. https://doi.org/10.5281/zenodo.23169743

Permanent link: 10.5281/zenodo.23169743. The paper describes the snapshot of 5 October 2026, deposited with it as JSON and CSV. The files on this site are rebuilt every day, so cite the deposit when the figures have to stay put, and link this site when you want what is open now.

No tracking cookies. No third-party analytics. No tracking pixels.

• We count page views to know whether anyone is using this. IP addresses are never stored: we keep a salted hash that changes every day, so it cannot be linked back to you or followed across days.
• Searching the calls asks for nothing. The keyword search, the filters, the deadline list, every call's own page with the funder's exclusion clauses quoted, and the bulk files: no email, no account, no condition. When a search returns nothing at all we keep the words typed and the date, so we know what the register is missing. That record holds no address and no identifier.
• A match asks for an address, because the report is the email. The institution match, the consortium match and the partner match each read a public research database, build a profile of what the organisation publishes on, and rank the register against it. The page tells you what was checked, what was set aside and why, and how much came back: the ranked report goes to your inbox. Without an address there is nowhere to send it, so the match cannot run.
• A few matches a day, per address. Each one reads a research database and sends an email, so an address can have 5 reports in a day across all three kinds of match. Almost nobody reaches that; if you do and you need more, write to us and we will sort it out.
• The first time an address is used here it gets one line asking it to confirm. Nothing on a form can prove an address belongs to the person typing it, and we will not send a report to an address somebody else typed in. An address that never confirms is asked twice in its life and then never again.
• Two records, deliberately kept apart. We keep which organisations were matched: the name typed, the country chosen, and whether we could answer, so we know who this is useful to and which matches we are failing. That record is about organisations and not about you: no address, no identifier, no session, and the date only. Your address sits in a separate delivery record that holds the report and a date, with no organisation name in it. We do not join the two, and on the pages that take an address we record no visitor identifier at all.
• How long. A delivery address goes when the report has gone, and within 30 days whatever happens: an address we are no longer using to send you something is an address we should not be holding. An address that never confirmed goes within seven days. The record of which organisations were matched contains no address, so it stays.
• Institution level only. Every match uses public bibliographic data aggregated at institution level. We do not build profiles of named researchers, and the register holds none.
• One email, and no list. The report is sent through Resend, our mail provider, which is a United States company: ask us and we will send you their terms. Your address is used to send that report. It is not an account, it is not added to the digest, and it is not passed to anyone. The report ends with one short paragraph about SNM's own paid service, which is how this is paid for, and you can reply to say stop.
• The digest is a separate form with its own opt-in. Double opt-in: one confirmation email, and nothing more unless you click it. One-click unsubscribe in every message. Nothing you type into a match form enrols you in it.
• There is exactly one cookie, and only if you deliberately create it: if you choose to use your own OpenAlex allowance, the key you paste is stored in a cookie on your own device so we can pass it to OpenAlex. It is never written to our database and never logged. Nothing else on this site sets a cookie, and you can remove that one from the same page.
• Your rights, and a mailbox that is read. Write to contact@s-n-m.eu for access, correction, erasure, restriction, portability or to object, and we answer inside a month. Nothing here makes an automated decision about a person: the matches rank organisations against calls. Full details, and who to complain to, are on the legal notice.

Institution profiles are built from OpenAlex (CC0). Controller: Scientific Network Management, S.L., Barcelona, Spain. Company details and contact.