UKRI AI, HPC and Quantum Computing Grant 2026: Up to £3.75 Million for Hybrid Workflows

UKRI AI, HPC and Quantum Computing Grant 2026: Up to £3.75 Million for Hybrid Workflows

UK Research and Innovation (UKRI) is opening a substantial grant call for teams working at the point where quantum computing meets artificial intelligence and high-performance computing. The opportunity is delivered by the Engineering and Physical Sciences Research Council (EPSRC) with the Science and Technology Facilities Council (STFC), and forms part of the UKRI Quantum Programme.

The call is not a general quantum research fund. Its focus is practical integration: applicants must show how quantum computing will operate as part of an end-to-end workflow that also uses classical high-performance computing, AI, machine learning, or a combination of these. The proposed work must be built around a clearly defined application and a named or otherwise defined end-user. UKRI is looking for demonstrators that can show meaningful advantages in areas such as time to solution, fidelity, cost, energy efficiency, or usability.

The opportunity opens on 4 September 2026 and closes at 4:00 p.m. UK time on 3 November 2026. Projects can run for up to three years and are expected to start on 1 June 2027. The full economic cost of a project can be up to £3.75 million, with EPSRC funding 80% of the FEC. This is a high-value, team-based research and innovation grant for established UK research environments, not an individual scholarship or a small startup award.

Key details

What the grant is designed to support

The call sits within Quantum Mission 1 of the UK’s National Quantum Strategy. UKRI describes three broad outcomes for that mission: a competitive UK quantum computing ecosystem, demonstrable economic and societal value, and widespread adoption and integration. For this grant, those outcomes translate into research that can move beyond an isolated algorithm or laboratory test and show how quantum resources can be used with existing computing systems for a valuable task.

Quantum computing must be central to the proposed research. An application that simply adds a quantum component to a conventional AI or HPC project is unlikely to fit. The central question should be what the hybrid system makes possible, and why the work cannot be answered as well by a purely classical workflow. The comparison does not need to claim that quantum hardware already wins every benchmark. It does need to define a credible route to testing, measurement, and improvement.

UKRI identifies several technical challenges that applicants may address. These include hardware descriptors that help automate orchestration decisions, performance models for resource allocation and expected runtime, workflow optimisation for data encoding, and seamless transpiling and error mitigation. Other possible themes include AI methods for quantum performance, software tools that connect hybrid systems to classical software, scheduling and instruction-specific tasking, architecture simulation, reproducible benchmarking, and machine learning methods for error suppression.

The call also points to application areas including chemistry, materials, optimisation, partial differential equations, digital twins, and AI for science. Its examples of real-world sectors include healthcare, energy, finance, transport, defence, manufacturing, logistics, and advanced materials. These examples are not a closed list. They are useful signals about the kind of problem statement the funder wants: a concrete need, a realistic user, and a measurable reason for combining quantum and classical resources.

Every funded project must develop, at a minimum, demonstrators showing where an end-to-end hybrid quantum-classical workflow offers meaningful advantages for selected use cases. The project should also map its outputs to Quantum Mission 1 milestones, including progress toward larger-scale quantum operations and future fault-tolerant applications.

Who is a good fit

The formal organisational requirement is important. The applicant must be based at a UK research organisation eligible for EPSRC funding. UKRI’s standard individual eligibility rules also apply. The opportunity is therefore most naturally suited to universities, research institutes, and other approved research organisations that can provide the necessary research environment and submit through their research office.

The technical team should be broader than a single quantum-computing laboratory. UKRI specifically anticipates cross-cutting teams involving some combination of physics, computer science, engineering, mathematics, software, hardware, and end-user expertise. Industry, government departments, non-departmental public bodies, and third-sector organisations can contribute where they are relevant to the use case and the intended impact.

An applicant may be named as project lead on no more than one application, although there is no limit on the number of applications on which a person can be a project co-lead. This makes internal coordination essential. A department should decide early which proposal each potential lead is actually backing, rather than discovering a lead conflict shortly before submission.

International researchers and organisations may still be involved, but UKRI says that the UKRI-RCN Money Follows Cooperation Agreement and the UKRI-IIASA agreement do not apply to this opportunity. A Norway-based or IIASA project co-lead cannot be included under those arrangements. International collaborators, and UK partners not based at approved organisations, should be included as project partners instead. Check the current UKRI rules and your organisation’s research office advice before finalising roles or costs.

The strongest fit is a team that already has three things: a technically credible hybrid-computing idea, access to the relevant computing resources, and an end-user who can define the problem and test whether the result matters. If your proposal has only one of these, it is probably not ready. A promising algorithm without a user is too abstract; a user need without a credible quantum role is too generic; and an ambitious workflow without secured access to hardware or simulation is too risky.

Funding, project length and required contribution

The project can have a full economic cost of up to £3.75 million. EPSRC will fund 80% of that FEC. This means the headline maximum is a project-cost ceiling, not a promise that every successful award will receive £3.75 million from EPSRC. Cost the work around what is necessary to answer the research question and build the demonstrator, then confirm the institutional contribution and eligible cost treatment with your finance and research offices.

The call expects funded projects to secure at least 1:1 additional funding from public or private sources across the project’s lifetime. UKRI says this additional funding should be clearly evidenced and will be revised as part of standard EPSRC monitoring and evaluation. This is not described as a requirement for matched funding from the host institution beyond the standard 20% FEC. Applicants should distinguish the two points: the usual FEC share is one issue, while the wider programme expectation about public or private funding is another.

Projects may last up to three years and have a fixed start date of 1 June 2027. A shorter project can be more persuasive if it has a focused demonstrator and a realistic evaluation plan. Do not fill three years with activities simply because the maximum is available. Explain what will be built, tested, measured, and handed on during the period, and show how the timetable reaches a useful result rather than ending with an open-ended research list.

Resources can include research expenses, travel, UKRI-funded research facilities, research technical support, research software engineers, data scientists, postdoctoral research assistants, consumables, and smaller equipment items costing less than £25,000 each. Funding can also support collaboration activities such as secondments, staff exchanges, and regular travel. The purchase of equipment over £25,000, including VAT, is not permitted under the call.

How to apply

Applications must be submitted through the new UKRI Funding Service. Je-S is not the submission route. The project lead completes the application, but UKRI expects team members and project partners to contribute. Only the lead research organisation can submit the application to UKRI.

The practical sequence is straightforward but should begin before the public opening date. First, confirm that the lead organisation and the proposed individuals satisfy UKRI and EPSRC eligibility rules. Next, agree the project lead, co-leads, specialists, partners, end-user, and institutional approvals. Then create or update the Funding Service account and select “Start application” when the call opens. The Funding Service allows applicants to answer questions directly, save work, and return to complete it. UKRI advises applicants to leave enough time to review the application in read-only form before sending it to the research office.

The application includes a 550-word summary intended to help UKRI identify appropriate assessors. That summary must make the AI, HPC, and quantum roles understandable to a broad reader. It should state the use case, the real-world problem, how the hybrid workflow is expected to help, and the connection to Quantum Mission 1. Avoid using the summary as a miniature technical paper full of unexplained abbreviations.

The Vision and Approach response must be supplied as one PDF of no more than six sides of A4, single-spaced, in 11-point Arial or an equivalent sans-serif font, with margins of at least 2 cm. An additional page for a diagrammatic workplan is allowed. The file name must use the unique Funding Service number followed by “Vision and Approach”, and the file must be no larger than 8 MB. If an attachment misses these requirements, UKRI says the application may be rejected.

The application also asks for core team details and a substantial Applicant and team capability to deliver response, with a stated 1,650-word limit. The assessment areas are vision and approach, applicant and team capability, resources and cost justification, ethics and responsible research and innovation, and user engagement. Treat each area as a decision the reviewers must make, not as a formality to complete after the science is written.

Timeline and planning checkpoints

The call opens on 4 September 2026 at 9:00 a.m. UK time. The deadline is 3 November 2026 at 4:00 p.m. UK time, and UKRI says it will not be possible to apply after that time. Your organisation may impose an earlier internal deadline for approvals, costing, partner checks, and final submission. Ask for that date immediately.

UKRI will hold a matchmaking workshop on 23 September 2026 for researchers, technology developers, policy experts, and end-users. Interested applicants can register by emailing [email protected]. The event may be useful for teams that have technical expertise but need a credible use-case partner or want to test whether a proposed application is understandable outside its home discipline.

Awards have a fixed start date of 1 June 2027 and can run for up to three years. Build backwards from that date when discussing recruitment, procurement, access to hardware, partner agreements, ethics review, data access, and demonstrator delivery. Do not assume that a successful award can simply begin when the team is ready; the published start date is fixed.

What to prepare before the application opens

Start with a one-page problem statement written for a technically literate non-specialist. Name the end-user, the operational or scientific decision they need to improve, the data or constraints involved, and the current classical baseline. Then explain where quantum computing enters the workflow and what the team will measure. If the benefit is expected to be lower energy use, faster optimisation, better fidelity, or improved usability, define the metric before discussing the method.

Prepare an evidence map for the four most important claims in the proposal: the problem is real, the end-user has a reason to participate, the team can run the proposed workflow, and the result could matter beyond a single experiment. Evidence might include prior results, benchmark data, user requirements, letters or statements from collaborators, secured facility access, software prototypes, or a documented route to obtain missing resources. The official page does not prescribe a particular letter format, so do not invent one; instead, make partner contributions concrete and check what your research office expects.

Resource planning should be equally specific. Separate staff effort, facility access, travel, collaboration, software, consumables, and small equipment. Explain why each major cost is needed for the demonstrator and how it connects to the workplan. If you need a major UKRI or European facility, contact that facility before applying to check feasibility and obtain a technical assessment if the Funding Service requires one.

Finally, prepare a risk register. Quantum hardware availability, cloud access, data quality, classical baselines, algorithm scaling, error mitigation, partner availability, and security or trusted-research concerns can all affect delivery. A credible proposal does not pretend those risks do not exist. It gives each important risk an owner, an early test, and a fallback that still produces useful evidence.

What reviewers are likely to look for

UKRI’s assessment criteria provide a clear guide. The vision should show excellent quality and importance, potential to advance understanding, timeliness, and impact on research, society, the economy, or the environment. It must also show significant progress toward Quantum Mission 1 and the case for convergence between quantum computing and AI or HPC.

The approach must be effective, feasible, and explicit about delivery risks. It should describe a transparent method where appropriate, explain how prior work will be built upon, and show how outputs will become outcomes. Reviewers will also look at the research environment, the team’s understanding of industry-led quantum computing, the primary customers and technology integrators, and how access to critical computing resources will be secured and managed.

User engagement has its own logic. The end-user should help inform requirements, validate approaches, and maximise impact. UKRI expects the team to use real-world data to gauge feasibility or to show how the workflow could enable something not possible at present. A partner whose name appears only in a support letter is weaker than a partner with a defined role in requirements, testing, feedback, and adoption decisions.

The capability section should demonstrate the right mix of skills rather than list every available specialist. Explain who owns the quantum work, who owns the AI or HPC integration, who manages the application domain, who will build and maintain software, who will handle governance, and who will manage the partnership. Make the handoffs visible. Hybrid projects often fail at interfaces, so a clear delivery structure is itself useful evidence.

Common mistakes to avoid

The first mistake is presenting a quantum method without a user-facing problem. This call is about integration and application-driven workflows, so a technical novelty claim alone is not enough. The second is describing a classical pipeline with quantum terminology added at the edges. Make quantum computing central and explain the interface between the components.

Another common problem is an unsupported advantage claim. Do not promise quantum superiority without a testable baseline, a metric, and a credible measurement plan. UKRI specifically mentions time to solution, fidelity, cost, energy efficiency, and usability. Choose the measures that fit your use case and say how the team will compare them fairly against classical, simulation, and AI-enabled baselines.

Avoid treating hardware access as a detail to solve after award. The Vision and Approach criteria ask how critical computing resources will be secured and managed, including contingency arrangements. The National Quantum Computing Centre can potentially support successful projects with hardware, simulation platforms, technical staff, workshops, and site visits, but collaboration with the NQCC is not mandatory. If your project needs that support, describe the route and status accurately rather than implying a confirmed commitment that does not exist.

Do not confuse the £3.75 million FEC ceiling with guaranteed cash, ignore the 20% standard FEC share, or omit the wider expectation of at least 1:1 additional public or private funding over the project lifetime. Do not budget prohibited equipment over £25,000. Do not submit after the deadline, overlook an earlier internal deadline, or put a Norway-based or IIASA co-lead into a role the call excludes.

Frequently asked questions

Can a startup apply by itself?

The published requirement is that applicants be based at a UK research organisation eligible for EPSRC funding. Startups and other companies may participate as partners where appropriate, but the official page does not say that a company can act as the lead applicant under this call. Confirm the proposed arrangement with UKRI and the lead organisation’s research office before relying on it.

Does the proposal need to combine all three areas equally?

No fixed equal split is stated. The proposal should support convergence of quantum computing with HPC, AI, including machine learning, or a combination. Quantum computing must be central, and the workflow should be end to end. The right balance depends on the problem and the evidence needed.

Is NQCC collaboration compulsory?

No. UKRI says collaboration with the National Quantum Computing Centre is not mandatory. NQCC may support successful projects with hardware, simulation platforms, technical staff, workshops, or visits, but applicants should describe any contact or access accurately.

Can the project start whenever the award is made?

No. Awards made through this opportunity have a fixed start date of 1 June 2027. The project can last up to three years.

Can a project lead submit several proposals?

No. A person may be named as project lead in no more than one application. UKRI says there is no limit to the number of applications on which a person can be a project co-lead.

Is matched funding required from the host university?

UKRI says there is no requirement for matched funding from the host institution beyond the standard 20% of FEC. Separately, funded projects are expected to secure at least 1:1 additional funding from public or private sources across their lifetime. Discuss both points with your research office.

Official links and next steps

Read the official UKRI funding opportunity page for the live application button, full criteria, Funding Service instructions, contact details, and any updates. Questions about the specific opportunity can be sent to [email protected]; technical questions about the Funding Service should go to the UKRI Funding Service helpdesk.

Before 4 September, identify the lead organisation, confirm the project-lead rule, contact the intended end-user, check access to hardware and simulation resources, and ask your research office for its internal deadline. Once the call opens, build the application around one defensible use case and one measurable hybrid workflow. Submit a complete, checked application by 4:00 p.m. UK time on 3 November 2026, and confirm the final live instructions immediately before sending it.