NSF Stakes $1.5B on X-Labs, an OT-Based Bet on Independent Research Teams

Academic PIs accustomed to NSF's standing grant programs now have a parallel track to weigh—a $1.5 billion, decade-long NSF X-Labs initiative whose first two topic solicitations close July 13 and July 24, 2026, according to NSF's May 14 announcement.

The program is not another standing solicitation. It is a procurement-style instrument that NSF is using to stand up independent, milestone-driven research teams outside the conventional university PI model. For investigators who have spent careers writing R01s and NSF CAREER awards, the structural details matter more than the headline.

What NSF actually committed to

The official NSF announcement puts the topline at $1.5 billion over ten years, released through what the agency calls an Other Transactions Agreement Solutions Offering. That mechanism—familiar in pockets of DoD, ARPA-H, and the Department of Energy, but new at this scale for NSF—lets the agency post topic announcements rather than full Federal Acquisition Regulation solicitations, and it lets it write Other Transaction contracts in place of grants or cooperative agreements.

Two topic announcements landed alongside the headline:

• Scientific Instrumentation for Sensing and Imaging. Written proposals due July 13, 2026, 5:00 p.m. Eastern. Scope includes quantum sensing, AI-driven computational imaging, and what NSF calls "entirely new chemical modalities."
• Quantum Systems: Interconnects and Integrated Photonics. Written proposals due July 24, 2026, with oral presentations the week of August 31–September 4, 2026. Scope is heterogeneous quantum system integration and components that transfer quantum information across distance.

NSF says the Solutions Offering will remain open through May 2028, with additional topics rolling out in the intervening months. The agency ran a public RFI in December 2025 to scope the first two topics; teams that responded then have a working sense of what reviewers want, which is a meaningful information asymmetry for anyone coming in cold.

Acting NSF Director Brian Stone framed the initiative around "independent, milestone-driven research teams," and White House OSTP Director Michael Kratsios called it "a bold step forward in revitalizing American innovation." Both statements track with the broader fiscal context: the President's FY2027 budget request proposed a 55% cut to NSF's core research and education accounts, a reduction that—if enacted—would dwarf X-Labs in dollar terms while shifting the agency's center of gravity toward fewer, larger, application-leaning programs.

The OTA pivot is the real story

For PIs whose mental model of NSF funding is built around the Proposal & Award Policies & Procedures Guide (PAPPG), the X-Labs mechanism is a different animal. Other Transaction authority bypasses the FAR and the standard NSF cooperative agreement, which has three immediate consequences.

First, the proposal is not a 15-page Project Description with five supplementary documents. The topic announcements call for written submissions structured around technical milestones, team composition, and a credible plan to move from concept to a deployable platform. Reviewers are looking for a venture-style narrative: what is the breakthrough, who is the team, what gets demonstrated by when, and at what cost. The closest analog in the federal portfolio is an ARPA-H Solutions Summit pitch—not a standing solicitation.

Second, awards are not made to an institution as the legal recipient with a PI as the responsible investigator. The mechanism funds the team directly, and NSF has been explicit that no traditional institutional affiliation is required. That does not mean university PIs are excluded—several of the strongest expected applicants are faculty who will spin out a separate entity or operate through a sponsored research model—but it does mean the conventional indirect cost recovery story does not apply. Whoever holds the OT contract negotiates costs against the milestone schedule, not against a federally negotiated rate.

Third, intellectual property terms are negotiated, not boilerplated. Bayh-Dole's default allocation does not automatically attach. PIs whose technology transfer offices are accustomed to running boilerplate term sheets should expect a different conversation about background IP, foreground IP, and government use rights before a proposal goes out the door.

Phase 0 looks small. Phase 1 changes the math.

The program structure is two-stage. Phase 0 awards are capped at $1.5 million per team across nine to twelve months and are designed for organizational formation—governance planning, milestone development, technical roadmapping, and the kind of partnership-stand-up work that is normally invisible to the federal funder. Teams that hit Phase 0 milestones can transition to Phase 1, which carries awards of up to $50 million per year over 24 to 36 months.

That ceiling is the headline. A successful Phase 1 award is on the order of $100 million to $150 million in committed funding under a single contract—comparable in scale to a mid-tier ARPA program, and well above what any single PI is likely to assemble through stacked R01s, NSF center awards, or DoD MURIs over the same horizon. The catch is that Phase 1 is conditional on Phase 0 execution, and NSF has signaled it will down-select aggressively. The expected funnel is dozens of Phase 0 awards across the first two topics, narrowing to a handful of Phase 1 transitions per topic.

For an academic group considering whether to compete, the calculus is not "should I write a proposal." It is whether the home institution can support a venture-style team structure for 9–12 months on a $1.5M Phase 0 contract, knowing the probability of a Phase 1 transition is single-digit and the IP terms are non-standard. That is a question for a department chair, a tech transfer office, and a research dean—not just a PI.

Why NIH-affiliated PIs should be reading this

X-Labs is an NSF program, but its mechanism choice matters for the broader NIH research community for two reasons.

First, ARPA-H has been operating under similar OT authority since its standup, and the model is migrating across agencies. NIA, NIAID, and NCI have all expanded their use of OT-style mechanisms over the last 18 months, often co-funded with ARPA-H. PIs whose portfolios straddle NIH disease areas and NSF instrumentation or quantum programs—neural imaging, single-molecule sensing, AI-driven diagnostics—are now seeing the same procurement-style instruments appear on both sides of the agency wall. The expertise required to write a competitive OT proposal is the same skill set, and it is not the skill set the NIH grant-writing apparatus has historically cultivated.

Second, the first X-Labs topics overlap directly with active NIH research priorities. Quantum sensing has obvious applications in MRI, MEG, and single-cell measurement. AI-driven computational imaging is at the core of NIH's BRAIN Initiative instrumentation push. "Entirely new chemical modalities" is broad enough to encompass biosensing platforms that would otherwise be funded through NIBIB or NIGMS. For a PI whose center of gravity is NIH but whose technical proposal could plausibly be framed as instrumentation, X-Labs offers a $50M/year ceiling that does not exist anywhere in the NIH portfolio.

The trade-off is the IP terms, the team-versus-institution structure, and the milestone discipline. These are real constraints. They are also negotiable in ways that an R01 budget is not.

What to do this week

Three concrete steps for any PI weighing a Phase 0 submission against the July 13 or July 24 deadlines:

1. Read both topic announcements end-to-end before drafting. They are short relative to a standard NSF solicitation, but the milestone language and the team-composition expectations are the parts reviewers will weight most heavily. The Quantum Systems topic in particular has an oral presentation round in late August, which means the written proposal is a gate, not the final evaluation.
2. Talk to your tech transfer office about OT IP terms now, not in week six. The single most common failure mode for first-time OT proposers is discovering at submission time that their institution has not negotiated background IP allocation. Two months is a tight runway. Six weeks is not.
3. Map the active solicitation landscape against your technical scope before committing to X-Labs as your primary 2026 submission. Quantum sensing, computational imaging, and integrated photonics each have parallel NSF, DOE, and ARPA-H opportunities open right now, and the OT mechanism's IP terms may or may not be the right fit for your team's commercialization plan.

Browse current quantum and instrumentation solicitations on Granted's grant search to compare X-Labs side-by-side with the standing NSF and DOE programs you would otherwise be writing into. Granted's editorial coverage tracks federal mechanism shifts as they happen, including the ARPA-H OT model that X-Labs is partially built from.

The $1.5 billion topline will get the press coverage. The mechanism is the part PIs will be living with for the next decade.