DARPA Priced Five of Seven Topics at Direct-to-Phase-II Scale. Release 6 Closes October 21, and the Whole Slate Is One Battlefield.

Most small businesses read a Department of War SBIR release as a menu: scan the topic list, find the one or two that brush against your technology, write to those. That habit is about to cost people money, for two reasons that converge in the next six weeks.

The first is that DARPA's FY26 Release 6 closes October 21, 2026 at 12:00 p.m. Eastern, and the slate is not a menu. It is a scenario. The second is that DoW FY-27 Release 1 pre-released October 7 with a rule change, effective October 1, 2026, that quietly ends the shotgun approach to defense SBIR.

Let us take them in order, because the strategic response to each is different.

Release 6: ten topics, and four of them are the same patient

DARPA's Release 6 pre-released September 2, 2026, opened September 23, and closes October 21. It carries seven SBIR topics and three STTR topics:

SBIR

• DPA26BZ06-DV023 — Novel Radio Frequency Sensing Technologies (Tactical Technology Office)
• DPA26BZ06-DV024 — New World Screwworm Networked Detection, Identification, and Surveillance System (Biological Technologies Office)
• DPA26BZ06-DV025 — Noninvasive Detection and Localization of Occult Hemorrhage (BTO)
• DPA26BZ06-DV026 — Influence Benchmarks for AI Systems (BTO)
• DPA26BZ06-DV027 — Casualty Operations and Resource Prediction Software, or CORPS (BTO)
• DPA26BZ06-DV028 — TRIAGE-X: Autonomous Casualty Triage and Treatment in Chemically Contaminated Mass Casualty Events (BTO)
• DPA26BZ06-DV029 — ICU-in-a-Box: Autonomous Extracorporeal Multiple-Organ Support Therapies (BTO)

STTR

• DPA26TZ06-DV004 — SHIELDER: Scalable Hard-Mask Materials with Improved Etch Resistance for Extreme-Aspect-Ratio Fabrication (Defense Sciences Office)
• DPA26TZ06-DV005 — Fuel-Flexible Spacecraft Electric Propulsion System (DSO)
• DPA26TZ06-DV006 — Localization, Characterization, and Modeling of Freestream Disturbances in Hypersonic Wind Tunnels (DSO)

Six of seven SBIR topics come from a single office. That alone is unusual. But look at what four of them do in sequence.

DV025 finds internal bleeding you cannot see. DV028 recognizes a chemical toxidrome and physically administers the antidote without a human in the loop. DV029 is a portable extracorporeal life-support platform — resuscitation plus cardiovascular and pulmonary support, with optional electrolyte and medication delivery — validated in large-animal models. DV027 predicts how many casualties are coming and what resources they will consume.

That is not four topics. That is one casualty, from point of injury through prolonged field care, in a contested environment where evacuation is not available and a physician is not present. DARPA has decomposed a single operational problem — mass casualty care without a hospital — into its component technology gaps and published them simultaneously.

The strategic consequence is specific: your proposal should name the adjacent topics. A triage autonomy proposal that does not acknowledge it will have to hand a patient to something like DV029, or consume a resource forecast like DV027, is describing a product that works alone. DARPA has just told you, in the structure of the release itself, that it is not buying things that work alone. Interoperability claims that would be speculative padding in a normal solicitation are, here, responsive to the actual portfolio.

The award sizes are the real signal

This is where most readers will mis-size their effort.

The standing assumption in defense SBIR is that Phase I is feasibility money: a few hundred thousand dollars, six to twelve months, prove the concept is not impossible. Across DoW, standard Phase I SBIR and STTR topics generally run from roughly $100,000 to $315,000, and Direct-to-Phase-II topics go up to $2 million.

Release 6 does not sit at the bottom of that band. Published topic values include:

• CORPS (DV027): $750,000
• TRIAGE-X (DV028): $2 million — structured as $1.5 million plus a $500,000 option
• ICU-in-a-Box (DV029): $2 million — same $1.5 million plus $500,000 option structure

A $2 million ceiling with an option tranche is Direct-to-Phase-II pricing. DARPA is not asking whether autonomous toxidrome recognition is conceivable. It is asking who can demonstrate it, and the ICU-in-a-Box topic says the quiet part out loud by specifying large-animal model testing in the topic itself.

Three things follow.

One: the maturity bar is a screen, not a preference. If your relevant work exists as a simulation, a literature review, and an enthusiastic deck, a $2 million demonstration topic with an animal-model deliverable is not a long shot — it is a mismatch. The money is the eligibility criterion.

Two: budget the option. A $1.5 million base with a $500,000 option means the government is reserving a decision point. Proposals that spend the full $2 million on a flat workplan misread the instrument. Structure the base period to end at a defensible, evaluable milestone, and put the work that only makes sense after that milestone into the option.

Three: animal-model and regulatory timelines do not compress. If DV029 is your topic and you do not already have an IACUC relationship and a large-animal facility arrangement, the realistic path is a teaming agreement signed before October 21 — not a plan to find one after award.

Note too that TRIAGE-X is administered through DARPA's Small Business Projects Office, not the technical office that generated the science. That is a contracting-pathway detail, and it matters for who answers your questions and how fast the award moves.

One hard constraint: the pre-release window for Release 6 — the only period in which you may contact Technical Points of Contact with topic-specific technical questions — ran September 2 through September 23 and is closed. For Release 6, you are writing without clarification. That is a reason to be conservative about inferring unstated requirements and generous about stating your assumptions explicitly in the proposal.

FY-27 Release 1: 25 topics, no DARPA, and a rule that changes tactics

While Release 6 closes, the next fiscal year's first release is already in pre-release.

DoW FY-27 Release 1 pre-released October 7, 2026, opens October 28, and closes November 25, 2026 at 12:00 p.m. Eastern. It carries 25 topics:

• Department of the Air Force: 16 (13 SBIR, 3 STTR)
• Navy: 6
• Office of the Secretary of War: 2
• Army: 1, tied to an xTech competition — and on a different clock: white papers due October 23, 2026, with finalist proposals not due until May 7, 2027

Phase I ceilings are up to $300,000 for Air Force and OSW topics and up to $315,000 for Navy topics including option. Direct-to-Phase-II topics run up to $2 million over 24 months.

Two observations worth acting on.

First, DARPA has no topics in FY-27 Release 1. If DARPA is your agency, your calendar for the next two months is Release 6 on October 21 and then a wait. Do not pad an Air Force topic because you want to stay busy in November; a thin response to a topic that is not yours is worse than a gap.

Second, and more consequential: effective October 1, 2026, DoW limits proposals on a per-topic basis. The tactic of submitting several variants against one topic to see which framing lands is over.

This is a direct downstream effect of the reauthorization. The Small Business Innovation and Economic Security Act of 2026 was signed April 13, 2026, extending SBIR and STTR through September 30, 2031, and it brought application caps, foreign-risk screening, a 90-day award clock, and the $30 million Phase II strategic breakthrough vehicle with it. We covered the statutory architecture in the reauthorization analysis. Per-topic limits are that statute arriving as a submission rule.

The practical effect is that internal selection now happens before submission rather than at the agency. Firms that used to resolve "which of our three angles is strongest?" by letting evaluators decide must now resolve it themselves — and the honest version of that exercise usually reveals that two of the three angles were never strong.

It also means the pre-release window becomes the highest-leverage two weeks in the cycle. For Release 1, direct TPOC contact is available October 7 through October 27 only. One well-aimed technical question in that window is now worth more than a second proposal, because the second proposal no longer exists as an option. Ask whether your interpretation of the threshold requirement is right. Ask what the government considers an acceptable demonstration environment. Then write once.

What to do in the next two weeks

If you are responding to Release 6 by October 21: confirm your SAM.gov registration is active and your DSIP submission is started today, not on the 20th. Size your effort to the published topic value — a $2 million option-structured topic expects a demonstration plan, a facility, and a team, and the TPOC window for asking about any of it has already closed. Name the adjacent topics in the slate your system would have to interoperate with.

If you are looking at FY-27 Release 1: use October 7–27 for TPOC contact, then pick one topic and one proposal per topic. Check whether your target is a standard Phase I or a Direct-to-Phase-II, because the ceilings differ by a factor of six and the evidence expectations differ by more than that.

And if you have not yet taken a Phase I, note that the on-ramp infrastructure moved too — SBA awarded its FY2026 FAST cohort on September 22, which determines whether free proposal assistance exists in your state. The commercialization standards that now govern Phase II progression also tightened this fall; see Granted News.

The fiscal year turned over. So did the rules about how many shots you get.