DARPA Made Its SBIR Schedule Predictable. Here's How to Build a Rolling Pipeline Around Release 5 and the First-Wednesday Calendar

For years, working with DARPA's SBIR program felt like waiting for weather. Topics dropped when they dropped, submission windows opened and slammed shut on their own schedule, and the small businesses that won were disproportionately the ones with someone whose full-time job was watching for the release. In FY26, DARPA quietly changed the physics of that game. New SBIR/STTR research topics now pre-release on the first Wednesday of every month, on a predictable, repeating cadence. Release 5 pre-releases August 5, opens for submission August 26, and closes September 23. Release 6 follows one month later — pre-releasing September 2, opening September 23, and closing October 21.

That regularity is easy to skim past as a scheduling footnote. It is not. A predictable release calendar is the single most valuable planning asset a resource-constrained small business can have, because it converts the hardest part of DARPA SBIR — being ready when a relevant topic appears — from luck into logistics. The companies that understand this are already restructuring how they staff, watch, and write. Here is what the new cadence actually is, and how to build a pipeline around it.

The mechanics of the first-Wednesday cadence

Each monthly release moves through three fixed stages, and the gaps between them are consistent enough to plan a quarter in advance:

• Pre-release (first Wednesday of the month). Topics become visible. Crucially, this is the only window in which small businesses may contact the Topic Authors (TPOCs) with technical questions. Once the topic officially opens, that channel closes.
• Open (roughly three weeks later). The submission portal accepts proposals. For Release 5, this is August 26.
• Close (about four weeks after open). The window shuts. Release 5 closes September 23 at noon Eastern, following the pattern set by Release 4, which closed August 19.

The load-bearing detail is the length of each stage. The pre-release-to-open gap is about three weeks; the open-to-close window is about four weeks. That means the total runway from first sighting to deadline is roughly seven weeks — but the portion during which you can talk to the government is only the first three. A company that waits until the topic opens to start writing has already forfeited the most valuable phase and compressed its usable runway by nearly half.

For a detailed look at how a single release plays out in practice, our analysis of DARPA's July 22 Release 4 topics — geopolitical forecasting, LLM-symbolic fusion, and radiation-hard memory, all closing August 19 — walks through why the four-week window is unforgiving and why pre-release engagement is decisive. Release 5 will follow the same rhythm with a new slate of topics.

Why predictability changes the strategy

Under the old ad-hoc model, the rational posture for a small business was reactive: monitor constantly, then scramble. Under a fixed monthly calendar, the rational posture flips to proactive capacity planning. Three things become possible that were not before.

You can pre-commit writing capacity. If you know a pre-release lands the first Wednesday of every month and the deadline falls roughly seven weeks later, you can block your proposal team's calendar for the year in advance. The bottleneck in SBIR is rarely ideas — it is having qualified writing and technical hours available in the exact four-week window when a fit appears. A predictable calendar lets you reserve those hours before you even know the topics.

You can build a topic-watching routine instead of a topic-watching person. The first Wednesday becomes a standing internal event: review the new topics, score each against your core capabilities in a single triage meeting, and decide go/no-go within 48 hours. That discipline matters because the pre-release contact window is short. A company that triages on Wednesday and reaches out to the TPOC on Thursday has days of advantage over one that stumbles onto the topic when it opens three weeks later.

You can run a rolling pipeline across releases. Because releases stack monthly, a company is never betting everything on one window. If Release 5 has no fit, Release 6 is only 30 days out and already on the calendar. This changes the economics of the no-bid decision: passing on a marginal topic is far easier when the next at-bat is guaranteed and dated, which in turn raises the quality of the proposals you do submit.

Building the pipeline: a concrete cadence

Translate the calendar into a repeatable operating rhythm:

1. First Wednesday — triage. Pull the pre-release topics the morning they drop. Score each against a short internal rubric: Does it map to an existing capability or prototype? Do we have past performance we can point to? Can we realistically staff a proposal in the next month? Kill the marginal ones fast.
2. First week — engage the TPOC. For every go, send substantive technical questions to the Topic Author during the pre-release window. This is the only sanctioned pre-proposal contact, and thoughtful questions both sharpen your approach and signal seriousness. Do not waste it on logistics you can find in the instructions.
3. Weeks two to three — draft against the open date. Use the pre-release-to-open gap to build the technical volume while the topic is still fresh and the TPOC dialogue is live. Aim to have a complete draft before the portal opens, not after.
4. Weeks four to seven — refine, cost, and submit early. Once open, the four-week window is for tightening, compliance review, and the commercialization and cost sections — not for starting. Submit with days to spare; the portal's closing hours are notoriously congested.

The eligibility and readiness fundamentals still apply

The predictable calendar lowers the timing barrier but not the substantive ones. DARPA SBIR remains open to U.S. small businesses meeting SBA size and ownership standards, with the usual majority-U.S.-ownership and employee-count thresholds; STTR additionally requires a partnering research institution. Registrations in SAM.gov, the SBIR company registry, and the submission system must be current before a window opens — a lapsed or in-process registration is one of the most common and most avoidable reasons a ready proposal cannot be filed. The strategic gift of a fixed calendar is that it removes any excuse for letting those registrations expire: you know exactly when you will need them.

It is also worth remembering the broader 2026 context. SBIR/STTR was reauthorized through 2031, proposal caps and foreign-risk screening have tightened, and a new $30 million Strategic Breakthrough Award ceiling has raised the top end of what the program can fund. DARPA's monthly cadence sits inside that larger, more defined structure — a program that is simultaneously more generous at the top, more scrutinized on foreign ties, and more predictable in its rhythm than it was two years ago.

The bottom line

DARPA did not announce its new SBIR calendar as a strategy shift, but that is what it is. By fixing pre-releases to the first Wednesday of each month and holding a consistent seven-week runway with a three-week contact window inside it, DARPA has handed small businesses the one thing they could never manufacture on their own: foresight. Release 5 pre-releases August 5, opens August 26, and closes September 23. Release 6 is already dated behind it. The companies that win the back half of FY26 will not be the ones with the best luck — they will be the ones who turned the calendar into an operating rhythm and were writing before their competitors knew a topic existed.