DARPA LUC (HR001126S0014): What Startups Need to Know
Below is a brief summary. Please check the full solicitation before applying (link in resources section).
Quick Answer
DARPA's Information Processing Techniques Office (IPTO) has released the Lightweight Universal Codec (LUC) Broad Agency Announcement, solicitation number HR001126S0014. LUC funds development of a universal, adaptive encoder-decoder that can send and receive information using any known error correction code, and can switch between codes on the fly without redesigning the hardware or software. The program is 39 months across two phases, with only one technical area open for performers. Abstracts are required and due July 22, 2026. Full proposals are by invitation only, following abstract review, with a submission closing date of August 28, 2026. Multiple awards are anticipated, using Procurement Contracts, Other Transaction Agreements, Cooperative Agreements, or Grants.
Program Overview
LUC ("luck") is built on top of DARPA's earlier GRAND (Guessing Random Additive Noise Decoding) research, which produced a universal decoder capable of decoding many different families of error-correcting codes using a single algorithm, rather than needing a separate decoder built for each code type. LUC's job is to build the other half of that pair: a universal encoder that can dynamically select among many error-correcting codes in real time, pair it with a GRAND-style universal decoder, and integrate the resulting codec into an existing communication system.
The practical problem DARPA is solving: today's communication systems lock in a specific error-correcting code (or a very small set of them) before a transmission ever happens, because dynamically evaluating and switching codes has historically been too computationally expensive. That means systems can't adapt in real time as channel conditions and noise environments change. LUC removes that constraint by making the code selection itself dynamic, informed by the decoder's own noise model and mission-specific optimization goals.
Importantly, LUC assumes the set of codes it adapts across is already known in advance. Inventing entirely new codes on the fly is explicitly out of scope. This program is about universal, adaptive switching among known codes, not code discovery.
DARPA is explicit that it wants revolutionary advances here, not evolutionary improvements to existing codec practice.
Who This Is For
LUC is a strong fit for organizations working in:
Forward error correction and channel coding
Communications hardware and software-defined radio
Decoder architectures, including GRAND-style universal decoding
Low-power, low-latency signal processing and ASIC/FPGA design
Adaptive and cognitive communication systems
Integration of novel codecs into existing transceiver platforms
This is a narrowly scoped, single-technical-area program. It is not a fit for teams proposing new code families themselves, systems that require classified handling (classified submissions won't even be accepted), or work centered on platform changes like moving from a software-defined codec to a chip-based codec within the same device, since that kind of comparison is explicitly excluded from the metrics.
Program Structure
LUC has one technical area, TA1, Codec Design and Integration. DARPA is only soliciting performers for TA1. Government test and evaluation (T&E) partners will support assessment of every candidate technology, but performers do not propose into a separate T&E technical area the way some other DARPA BAAs are structured.
TA1 performers are asked to do three things:
Build a universal encoder that can adaptively select among multiple classes of error-correcting codes in real time
Integrate that encoder with a universal decoder, such as a GRAND implementation, to form a complete universal codec
Integrate the resulting codec into an existing, performer-nominated communication system
The program runs 39 months across two phases:
Phase 1, Base (12 months): Codec and system design, modeling, and simulation. DARPA and its T&E partners formally assess progress twice, once at month 6 with an initial codec design review and again at month 11 with an end-of-phase evaluation against the Phase 1 metrics. Performers who submit all deliverables and successfully meet every criterion may be advanced early into Phase 2.
Phase 2, Optional (27 months): Integration, field testing, and hardware delivery. Performers whose option is exercised integrate their codec into transceivers and run field tests at multiple locations, with one field test and one design review occurring in year two and again in year three. Each performer must deliver at least two LUC-enhanced communication devices and two unmodified "stock" devices to the government T&E team for A/B testing on size, weight, and power. LUC-enhanced devices must run at full rate in real time on fixed codes, meaning they can't be slower than a stock device except during the moment of an actual code switch.
Program Metrics
DARPA will score TA1 performers against six metrics, each with separate targets for Phase 1 and Phase 2, measured relative to the performer's own selected technology platform:
Codes supported on a single algorithm: Phase 1 requires catalogued Polar Codes for control channels per 3GPP TS 38.212 v19.3.0, five representative Reed-Solomon or Turbo codes, and one example of Modular Polar Long Codes. Phase 2 expands this to two examples of Advanced Modular Polar Long Codes, one representative AI-enabling 3GPP LDPC code, and 100 additional codes supplied by mission partners.
Channel latency: 25 percent reduction versus state of the art in Phase 1, 50 percent reduction in Phase 2, tested across multiple error-laden environments.
Signal gain in dynamic noise environments: 1 dB gain over state of the art in Phase 1, 6 dB gain in Phase 2.
Block Error Rate at Eb/No of 5 dB: 1 percent in Phase 1, 0.1 percent in Phase 2.
Receiver decoder power requirements: 10 times better than state of the art in Phase 1, 15 times better in Phase 2.
Time to switch code classes: Within 1 second after the initial complete communications frame in Phase 1, dropping to within 100 milliseconds in Phase 2.
Funding Structure
LUC does not publish a specific dollar ceiling or per-award funding range in the BAA. What is specified:
Multiple awards are anticipated
Award instruments may include Procurement Contracts, Research Other Transaction Agreements (10 U.S.C. Section 4021), OT for Prototype Agreements (10 U.S.C. Section 4022), Cooperative Agreements, and Grants
Resource sharing may be required for Research OT or OT for Prototype Agreements
DARPA explicitly discourages budgets that aren't tightly calibrated to the proposed technical approach. Submissions with a budget number that isn't backed by clear evidence connecting the level of effort to the cost will be reflected negatively in the evaluation
The IP and data rights section of this BAA is unusually detailed and worth understanding before you propose. DARPA is explicit that this solicitation is for research and experimentation only, not for production, procurement, fielding, or commercialization of any resulting technology. Any prototypes or test articles built under LUC are test articles only, not end items, and the government does not intend to rely on any LUC output for operational use. If DARPA later wants to move a LUC technology toward production or deployment, that requires a completely separate agreement with independently negotiated terms. Proposers are also required to submit an IP Management Plan and a Software Bill of Materials with their proposal, disclosing background IP, known third-party patents, licensing needs, and mitigation strategies, and performers carry Freedom-to-Operate responsibility and patent indemnification obligations for their own technology.
Key Dates
Posting Date: July 6, 2026
Proposers Day: July 8, 2026
Question Submittal Closes: July 17, 2026, 2:00 PM ET
Abstract Due Date: July 22, 2026, 2:00 PM ET (required)
Proposal Submission Closing Date: By government invitation only, August 28, 2026, 5:00 PM ET
Oral Presentations: By government invitation only, estimated to begin September 1, 2026, with the exact date confirmed in writing by DARPA
Unlike a lot of DARPA BAAs, an abstract here is not optional. Every proposer must submit one, and only proposers who receive an Invitation to Proceed are permitted to submit a full proposal or participate in oral presentations. There is no path to a full proposal without first clearing the abstract stage.
Eligibility
Open to all responsible U.S. and non-U.S. sources, including small businesses, Historically Black Colleges and Universities, and minority institutions, though there is no funding set-aside reserved for any of these categories. Non-U.S. participants must comply with applicable nondisclosure agreements, security regulations, and export control laws.
FFRDCs, UARCs, and government entities including national labs face significant restrictions here. DARPA will not award new contracts to these entities as prime performers under this solicitation, and any direct prime proposal from one of them may be deemed non-conforming. They can potentially participate as subcontractors, but only if their role and point of contact are clearly defined in the technical proposal, their costs are excluded from the cost proposal (a rough order of magnitude estimate in the technical proposal only), and, if UARCs are involved, an organizational conflict of interest mitigation plan is included. Any of these entities that do get funded will be required to share their work with other performers on the same program.
Organizations cannot simultaneously act as a technical performer and provide Systems Engineering Technical Assistance, Advisory and Assistance Services, or similar support services on this program, unless the DARPA Deputy Director grants a written waiver.
Security and Compliance Requirements
LUC work is expected at both the unclassified and CUI levels. Classified abstracts or proposals will not be accepted under any circumstances.
CMMC Level 2 is required for this entire solicitation, which is a notably higher bar than a lot of other DARPA programs that only require Level 1 at submission. To even be eligible for award, the proposer's CAGE code must already have a valid, unexpired Final Level 2 CMMC status for TA1 recorded in the Supplier Performance Risk System (SPRS), along with a current annual CMMC affirmation from an authorized Affirming Official. This needs to be in place before award, not something you can acquire afterward. CMMC requirements flow down to any subcontractor touching FCI or CUI. Awards made after November 10, 2026 additionally require a CMMC Third-Party Assessment posted in SPRS.
Because some LUC research may disclose performance characteristics of systems or manufacturing technologies unique and critical to defense, some awards will carry publication restrictions requiring DARPA's permission before releasing results, even though DoW policy generally favors unrestricted publication of fundamental research.
How the Selection Process Works
LUC uses a two-stage gate that's stricter than most DARPA BAAs. You cannot skip straight to a full proposal.
Stage 1, Abstract (required): Evaluated against four criteria in descending order of importance: Technical Comprehension (does the proposed universal encoding-decoding approach make sense, including channel inference, code selection, code convergence, and a concrete integration and backward-compatibility plan), Technical Ability (does the team have verifiable, demonstrated expertise in error-correcting codes, decoders, and communications systems), IP Considerations (are data, software, and patent rights clearly identified along with third-party risk and mitigation), and Cost Considerations (is the proposed level of effort logically tied to the resource ask).
Every proposer gets one of two outcomes: an Invitation to Proceed, or a non-selection notice. Non-selected proposers cannot submit a full proposal or participate in oral presentations under this BAA. An Invitation to Proceed does not guarantee an eventual award.
Stage 2, Full Proposal and Oral Presentation: Only for proposers who received an Invitation to Proceed. The full proposal package (slide deck, cost proposal spreadsheet and narrative, FRRBS Common Forms, and a redlined sample agreement if requesting an OT) is submitted and then presented in person at DARPA's Arlington, VA offices. Each oral presentation session runs 1 hour and 15 minutes: 5 minutes of DARPA opening remarks, a 30-minute presentation, a 10-minute break, and a 30-minute Q&A. DARPA records every session. Presentations should be technical, not promotional, and DARPA expects the actual technical personnel who can field detailed questions to be in the room.
Full proposals and oral presentations are then scored against four criteria in descending order of importance: Overall Scientific and Technical Merit, Proposer's Capabilities or Related Experience (including the ability to meet the CMMC Level 2 requirement), Potential Contribution and Relevance to the DARPA Mission (including how well you've identified and mitigated patent and Freedom-to-Operate risk), and Schedule and Cost Realism.
What You Need to Submit
Abstract (required): Submitted through the DARPA BAA Tool (BAAT) using the required LUC Abstract Template, at the unclassified or CUI level.
Full Proposal (invitation only): Details are provided in your specific Invitation to Proceed letter, but proposers should expect to need a presentation slide deck in English, a cost proposal spreadsheet and narrative, the FRRBS Common Forms (Biographical Sketch and Current and Pending Support), and, if requesting an OT award, a redlined sample agreement showing your proposed negotiated terms.
Program Meetings and Reporting Cadence
Kickoffs alternate between Washington, DC and San Diego, CA, starting in Washington, DC, and run one day
Field tests run five to seven days
The Phase 1 end-of-phase evaluation runs three days; the Phase 2 end-of-phase evaluation runs five days
PI meetings and design reviews are conducted virtually
Performers must meet with DARPA virtually no less than once every two weeks (roughly 30 minutes each) to report technical and administrative progress
DARPA may conduct unannounced site visits to performer facilities at its discretion
Required deliverables include technical documents, prototype hardware with full installation and diagnostic documentation, commented source code and build scripts, annotated slide decks within one week of every program event, quarterly technical status reports within 15 days of quarter-end, monthly financial status reports within 15 days of month-end, a post-exercise report within 30 days of every field test and end-of-phase evaluation, a final phase report within 30 days of the end of each phase, and a final report at the close of the program.
Why This Matters Beyond the Award
Every modern communication system, from 5G handsets to satellite links to tactical radios, depends on error correction, and right now nearly all of them are locked into a fixed code chosen at design time. A universal, adaptive codec that can switch codes in real time based on actual channel conditions is directly relevant to commercial wireless, satellite communications, and IoT, not just defense radios. A company that builds real low-power, low-latency universal codec technology under LUC is building IP with a genuinely broad commercial runway, on top of getting exposure to DARPA's GRAND research base.
Frequently Asked Questions
What is DARPA LUC? LUC, or Lightweight Universal Codec, is a DARPA IPTO program funding the development of a universal, adaptive encoder-decoder that can send and receive information using any known error correction code, and dynamically switch between codes in real time without a hardware or software redesign.
What is the solicitation number for LUC? HR001126S0014.
When are LUC abstracts due? July 22, 2026, at 2:00 PM Eastern Time. Unlike many DARPA BAAs, submitting an abstract is required, not optional, and you cannot submit a full proposal without first receiving an Invitation to Proceed based on your abstract.
When are full LUC proposals due? Full proposals are by government invitation only. The proposal submission closing date is August 28, 2026, at 5:00 PM Eastern Time, but only for proposers who already received an Invitation to Proceed.
How long is the LUC program? 39 months across two phases: a 12-month base Phase 1 focused on design, modeling, and simulation, and a 27-month optional Phase 2 focused on integration and field testing.
How much funding does LUC provide? DARPA has not published a specific dollar ceiling in the BAA. Multiple awards are anticipated, and funding instruments may include Procurement Contracts, Research Other Transaction Agreements, OT for Prototype Agreements, Cooperative Agreements, and Grants. DARPA explicitly discourages budgets that aren't tightly tied to the proposed technical level of effort.
What technical area does LUC solicit performers for? Only one: TA1, Codec Design and Integration. Performers build a universal adaptive encoder, integrate it with a universal decoder such as a GRAND implementation, and integrate the resulting codec into an existing communication system.
Does LUC allow proposers to invent new error-correcting codes? No. LUC assumes the set of codes a codec will adapt across is known in advance. Generating novel codes on the fly is explicitly out of scope.
What CMMC level does LUC require? CMMC Level 2, which is required for the entire solicitation. Proposers must already have a valid, unexpired Final Level 2 CMMC status for TA1 recorded in the Supplier Performance Risk System, along with a current annual CMMC affirmation, before they are eligible for award.
Can classified work be proposed under LUC? No. LUC anticipates work at the unclassified and CUI levels only. Classified abstract or proposal submissions will not be accepted.
Does DARPA intend to produce or field LUC technology directly from this program? No. DARPA is explicit that this solicitation is for research and experimentation only. Any prototypes are test articles, not production end items, and the government does not intend to rely on LUC results for operational deployment. Any future production or fielding would require a completely separate agreement with independently negotiated terms.
Are FFRDCs, UARCs, and national labs eligible to propose on LUC? Not as prime performers. DARPA will not award new contracts to these entities directly under this solicitation, and a direct prime proposal from one may be deemed non-conforming. They may participate as subcontractors only under specific conditions, including a clearly defined role, excluded funding from the cost proposal, and, for UARCs, an organizational conflict of interest mitigation plan.
What happens if my LUC abstract isn't selected? You'll receive a non-selection notification and will not be permitted to submit a full proposal or participate in oral presentations under this BAA. There is no appeal path described in the solicitation.
