NSF Grants 2026: Programs, Deadlines, and PAPPG

Full answer: This guide is for researchers, research administrators, educators, graduate applicants, and deep-tech teams comparing NSF grants in 2026. NSF announced 12 new foundational-research opportunities in August 2026 providing over $1.5 billion, while its funding page lists the newly posted GRFP with an October 19, 2026 next required due date. Start with the applicant route and current opportunity, then apply PAPPG 24-1 plus Supplements 1 and 2. Don't choose a budget, deadline, or success estimate from an agency-wide average.

NSF Grants 2026: Current Programs, PAPPG Rules, and Route Choice

Data note: Grantsights reviewed the NSF funding portal, August 2026 foundational-research announcement, current PAPPG page, proposal-preparation guidance, merit-review guidance, and Award Search on September 10, 2026. Program status, dates, eligibility, amounts, and required documents belong to the exact current opportunity.

NSF says it offers hundreds of funding opportunities across research and education. That scale makes a generic list less useful than a route decision. A faculty research proposal, graduate fellowship application, undergraduate training site, shared-instrument request, rapid-response idea, and startup technology project do not use one common deadline or budget rule.

This page answers three questions in order: what is current, which NSF route owns the project, and which rules control the proposal. It also removes a common trap from broad NSF searches: a historical agency average does not tell an applicant what to request or how likely one program is to fund the work.

Last updated: September 10, 2026.

2026 NSF Funding Status

According to NSF, 12 foundational-research notices announced in August 2026 provide over $1.5 billion. The group covers one biological-sciences route, one computing route, four engineering routes, one geosciences route, and five mathematical and physical sciences routes.

Grantsights counted the 12 linked opportunities on NSF's August 2026 foundational-research page and grouped them by the directorate shown in each title. This is an announcement-level count, not 12 identical grants. The underlying opportunities have different project types, limits, deadlines, and review details.

The current NSF funding portal also lists newer opportunities outside that 12-notice group. For example, NSF shows the Graduate Research Fellowship Program as posted August 25, 2026, with an October 19, 2026 next required due date. Condensed Matter and Materials Theory was posted August 19, and two mathematical-sciences group and training routes were posted August 14. A posting date isn't necessarily a deadline, so open the exact record.

Use the NSF grant hub for a Grantsights route overview and NSF's Funding Search for current agency records. Do not reuse the date from a sibling program.

Pick the NSF Route Before Writing

Seven common applicant jobs map to different NSF routes. The route is the first decision because it changes who applies, which source controls, what is submitted, and when.

The decision is not always “which acronym sounds closest?” NSF's division and program information explains that division and program codes route the application electronically to the program officer. That makes program ownership an application field, not only a research-discovery preference.

Based on our analysis, a strong route screen has six fields: applicant type, project job, directorate, division, current opportunity, and recent award pattern. If one of those fields is unknown, the next action is source research, not prose drafting.

The Current PAPPG Rule Stack

Three policy units matter in 2026. According to NSF, PAPPG 24-1 remains the current base version and applies to proposals submitted or due on or after May 20, 2024. Supplement 1, NSF 26-200, applies to financial assistance awarded on or after December 8, 2025. Supplement 2, NSF 26-202, applies to financial assistance awarded on or after January 22, 2026.

NSF's proposal-preparation page says a program solicitation or broad agency announcement may modify the standard PAPPG instructions. That means a generic NSF checklist is never the final checklist. Reconcile the sources for the exact route and record which rule wins each difference.

Don't claim a draft guidance document is already effective. NSF's current PAPPG page can show proposed or replacement guidance alongside the current rule stack. Label consultation material, current policy, and opportunity-specific requirements separately.

Eight Common NSF Proposal Parts

NSF identifies eight common parts of a research proposal, while warning that the list is not exhaustive and other proposal types can differ.

1. A 1-page Project Summary with an overview, Intellectual Merit, and Broader Impacts.
2. A Project Description, typically up to 15 pages.
3. References Cited.
4. Senior or key personnel documents.
5. A proposal budget and justification by year.
6. Facilities, Equipment and Other Resources.
7. A 1-page mentoring plan when applicable.
8. A Data Management and Sharing Plan, up to 2 pages.

According to NSF's preparation guidance, the budget justification is typically up to 5 pages and must document and justify each budget line. “Typically” matters: the exact opportunity can change the standard limit or require added material.

A useful compliance matrix has one row per requirement and these columns:

This matrix does more work than an old 15-page template. It catches route-specific changes and assigns the work to a person who can close it.

Intellectual Merit and Broader Impacts

Two National Science Board-approved criteria anchor NSF merit review: Intellectual Merit and Broader Impacts. According to NSF, reviewers consider both for each proposal, plus any program-specific criteria in the controlling opportunity.

Intellectual Merit asks about the potential to advance knowledge. Broader Impacts asks about the potential to benefit society and contribute to desired societal outcomes. The proposal should connect claims to a credible activity, responsible owner, resources, and a way to assess progress.

An applicant should not assume one criterion has a universal numerical percentage. Program context matters, and added criteria can matter. The practical review table is:

Use the NSF Intellectual Merit guide, Broader Impacts guide, and how NSF review works after the route decision. Those guides go deeper on the review criteria; this one covers the agency-level route.

What NSF Doesn't Tell You: The 12-Notice Map Still Needs a Program Screen

NSF's August 2026 page makes a large current portfolio visible, but a $1.5 billion announcement total does not give one applicant a budget target or odds estimate.

Grantsights classified the 12 notices into five decision groups: broad core research, large computing agenda, engineering divisions, geosciences core research, and mathematical or physical sciences programs. The classification changes the next action. A PI should move from the group to its exact opportunity, then to the division and program code, rather than searching for “average NSF grant.”

For example, one university in Ohio studying wave-based computing might see both a general electrical and computing systems route and the EFRI Wave-Based Computing route. The right choice depends on the agenda, team structure, scale, solicitation language, and program fit. The $1.5 billion group total cannot decide that.

The best bet is the route that matches the project's method, scale, team, and review community, even when a broader program appears easier to explain. Choosing the familiar acronym over the evidence is a weak fit.

Our route-evidence method is reproducible:

1. Record the exact opportunity URL and current status.
2. Record applicant and PI eligibility.
3. Identify directorate, division, and program code.
4. Extract the permitted proposal type, amount, duration, deadline, and preliminary gate.
5. Search NSF Award Search for recent related awards.
6. Compare abstracts for method, field, scale, institution type, and stated program.
7. Write a one-paragraph fit case and one-paragraph wrong-route case.
8. Contact the named program officer when the official route permits or encourages it.

This is public Grantsights information gain: the official pages supply records, while the method turns those records into an apply, reroute, or wait decision.

Budget and Award Data Without the Average-Award Trap

One agency-wide average cannot be applied across hundreds of opportunities. Fellowships, standard research awards, infrastructure proposals, cooperative agreements, rapid-response proposals, and small-business awards use different source units.

Before choosing an amount, record:

• the exact opportunity's ceiling, floor, range, or budget instruction;
• whether the amount is annual, total, direct cost, or total cost;
• project duration;
• indirect-cost treatment;
• participant-support treatment;
• subaward and consultant needs;
• equipment or facility costs;
• person-month commitments; and
• any cost sharing or prohibition on voluntary committed cost sharing.

Use NSF Award Search as context, not a substitute for the current solicitation. An award search result can show funded amount, abstract, program, recipient, and date, but it doesn't show unfunded applications and cannot produce a success rate by itself.

If a page or spreadsheet quotes an “NSF funding rate,” ask for the fiscal year, numerator, denominator, directorate, division, program, and proposal-action type. Agency, directorate, and program rates are different measures. Use the NSF grant success-rates guide for that measurement problem.

Find and Validate a Current Opportunity

A six-step search keeps current status and route ownership attached to the proposal.

Step 1: Start with the applicant. Separate faculty or institutional proposals, individual fellowships, undergraduate routes, and small-business paths.

Step 2: Define the project job. Is the work core research, research training, infrastructure, fellowship support, urgent research, early-concept work, translation, or commercialization?

Step 3: Search the current portal. Use NSF Funding Search and the relevant directorate page. Filter with the scientific method and community, not only a broad subject.

Step 4: Open the exact opportunity. Confirm status, dates, eligible organizations and PIs, award guidance, proposal type, required preliminary step, and contacts.

Step 5: Check award history. Compare recent awards from the same program. Do not mix records from a similarly named but different program.

Step 6: Build the source packet. Save the opportunity, PAPPG sections, supplements, preparation checklist, program contact, and award-search method with access dates.

A bad fit is a project that requires the reviewer to infer why it belongs in the program. Another red flag is a team that starts its 15-page narrative before confirming applicant eligibility or a required concept outline, pitch, letter, or internal limited-submission process.

Submission and Timing Controls

NSF's funding portal links to Research.gov and Grants.gov, but the exact opportunity decides the accepted route and requirements. Allow institutional processing time before the sponsor deadline. A proposal is not submitted merely because the PI uploaded a PDF.

Use this timing record:

Don't promise a universal number of months from submission to award. Review and decision timing can differ by route, panel schedule, proposal type, budget, and agency operations. If timing affects a hiring or fieldwork decision, ask the program contact and build a contingency plan.

Common NSF Grant Mistakes

Using a broad agency total as an applicant amount. The August 2026 $1.5 billion figure covers 12 opportunities. It is not an individual award range.

Copying a sibling deadline. GRFP's October 19, 2026 next required date does not belong to CAREER, REU, or a disciplinary core program.

Treating PAPPG 24-1 as the only rule. The current supplements and exact opportunity can change the controlling instruction.

Inventing reviewer experience. A source-backed guide should not claim access to private panel or proposal records that Grantsights cannot document publicly.

Turning award records into a success rate. Award Search does not contain the full application denominator needed for that calculation.

Starting with the narrative. Eligibility, route, program ownership, and preliminary gates come first.

Writing one generic Broader Impacts paragraph. Tie the outcome to activities, owners, resources, access, and assessment.

Assuming all NSF support is an institutional grant. GRFP is an individual fellowship route, while REU, CAREER, core research, and SBIR/STTR have different applicants and jobs.

Sources

Source review by Grantsights

We reviewed these sources on September 10, 2026. The exact opportunity owns its current status, deadline, eligibility, amount, and application instructions. The PAPPG and supplements own agency-wide policy. Award Search owns public award records, not application counts or individual funding odds.

The maintenance rule is to recheck the funding portal, exact opportunity, PAPPG page, and supplements whenever NSF posts a new route or changes the proposal-policy stack. Preserve dated portfolio facts as history, but never let them overwrite a newer opportunity.

What to Do Next

Start with the seven-row route table, not a blank Project Description. Choose the applicant lane, project job, directorate, division, current opportunity, and recent award comparison. Then build the compliance matrix and apply the two merit criteria to the exact work.

Use the Grantsights NSF hub, grant database, and report-ready award records to continue the search. If the route is CAREER, GRFP, REU, RAPID/EAGER, MRI, or SBIR/STTR, move to that canonical guide instead of stretching this broad page across a specialist decision.

The best broad NSF guide doesn't pretend one number describes the agency. It helps the reader reach the right current source before proposal effort becomes sunk cost.