British AI firms have simply had their greatest six months on report. Evaluation from HSBC Innovation Banking UK and Dealroom places UK AI funding at $12.6bn within the first half of 2026, greater than 4 instances the identical interval a yr earlier and near three-quarters of all enterprise capital invested within the nation. Whole UK enterprise funding reached $17bn, the strongest opening to a yr since 2022, and the UK took 39% of all European enterprise capital.
That capital is being deployed into companies whose unit economics look nothing just like the software program companies that skilled a technology of traders. The hole just isn’t small; it’s not non permanent, and most of it’s compute.
The quantity that broke the mannequin
ICONIQ’s State of AI survey, protecting roughly 300 executives at firms constructing AI merchandise, has AI product gross margins averaging round 52% in 2026. That’s up from 41% in 2024 and 45% in 2025, so the course is encouraging. It additionally sits 25 to 30 factors under the 75% to 85% that conventional software program has delivered for twenty years.
The distribution issues greater than the typical. Bessemer Enterprise Companions’ State of AI work discovered the fastest-scaling firms, these reaching $100m in annual recurring income in round eighteen months, had been working at roughly 25% gross margins, successfully shopping for distribution with compute. The extra capital-efficient cohort in the identical evaluation averaged round 60%. Each figures are a good distance from what a software program investor’s mannequin assumes.
Then there may be the discovering that ought to fear anybody underwriting these companies on a software program template. In ICONIQ’s knowledge, mannequin inference rises from 20% to 23% of whole spend as merchandise mature, whereas expertise prices fall from 32% to 26%. Learn that once more. The compute line grows as a share of spend as the corporate scales, moderately than shrinking.
Software program economics rested on the other assumption. Price of products offered was largely fastened, so every extra buyer arrived at near zero marginal value and margins expanded with quantity. In an AI product, the dominant value is incurred per question. It scales with utilization, typically sooner than utilization, and it doesn’t amortise. A enterprise that’s underwater at 100 clients doesn’t get rescued by reaching a thousand.
This flows straight via to how these firms are valued. The Rule of 40, which asks that progress plus profitability clears 40, quietly assumed COGS was principally fastened. Strip 25 factors out of gross margin and progress has to hold a far heavier share of the load to clear the identical bar. Buyers are adjusting, however the benchmarks in huge circulation nonetheless largely date from a unique value construction.
From an IT determination to a capital allocation determination
The consequence for founders and boards is that the place compute bodily runs has stopped being an engineering desire and grow to be a query in regards to the form of the P&L.
For many firms more often than not, the reply remains to be to name an API. In case your volumes are modest, your utilization is spiky, or you might be nonetheless altering the product each fortnight, paying a supplier for elasticity is straightforwardly appropriate. You’re shopping for optionality, and optionality is price paying for when you nonetheless want it.
The query arises on the level the place an organization is working its personal fashions, by itself GPUs, at sustained and predictable utilisation. That could be a actual threshold and a rising variety of British AI firms are crossing it, significantly these coaching or fine-tuning their very own fashions moderately than wrapping another person’s.
Right here I’ve to be straight in regards to the state of the proof, as a result of the web is filled with assured solutions and most of them are nugatory. Seek for the break-even level at which self-hosting beats renting, and you can see figures starting from two million tokens a day to 100 million tokens a month, from 8,000 conversations a day to eighteen months of near-total utilisation. They disagree by orders of magnitude, and the overwhelming majority are revealed by firms that promote one facet of the commerce. Anybody quoting you a common crossover quantity is promoting one thing. The trustworthy reply is that it will depend on your utilisation, your mannequin measurement, your site visitors profile and your engineering capability, and the one dependable calculation is the one performed by yourself invoice.
The case that’s truly documented
One firm has revealed its numbers in sufficient element to be helpful, and it’s price inspecting exactly as a result of it’s not an AI enterprise.
37signals, which makes Basecamp and HEY, spent $3,201,564 on cloud companies in 2022 and determined to go away. The Register reported that the corporate purchased round $700,000 of Dell {hardware} to exchange its cloud cases, and that this outlay was completely recouped throughout 2023 as contract commitments expired. Writing on his personal weblog, co-founder David Heinemeier Hansson put the projected saving at well over $10m across five years, with the {hardware} working in colocation moderately than in a facility the corporate constructed itself.
Three caveats belong with these figures, and Hansson provides most of them himself. The comparability is rarely totally like for like. An organization with no present racks pays to hire them, although he argues that value is trivial subsequent to a cloud invoice. The financial savings goal retains shifting because the enterprise grows and desires extra {hardware}. And Data Centre Dynamics, reporting the identical story, famous that the figures exclude {hardware} refresh cycles and the extra operations workers that proudly owning infrastructure requires.
The essential caveat is the one that the majority retellings skip: this can be a steady, predictable SaaS workload, not AI inference. It doesn’t switch routinely. What transfers is the precept, which Hansson states extra bluntly than any analyst would: as soon as the payments grow to be substantial, you owe it to your self and your traders to at the least do the maths.
That precept applies with extra drive to AI, not much less, as a result of the compute line is bigger and since ICONIQ’s knowledge says it grows moderately than shrinks with scale.
Why the British model has a lead time
There’s a wrinkle particular to firms working right here, and it’s one that doesn’t seem in any American evaluation of this query.
If the maths does level in direction of working your personal {hardware}, the capability to place it in is rationed. Britain’s demand-connection queue for the transmission community reached 125GW towards a nationwide peak demand of round 45GW, and the Nationwide Vitality System Operator pushed via its largest ever set of connection reforms in December 2025 to clear speculative tasks out of it. Excessive-voltage transformers carry lead instances of two to 4 years. Information centre emptiness throughout the UK has fallen from 27% in 2016 to eight% within the first quarter of 2026.
The sensible implication for a board is that this determination can’t be left till the compute invoice turns into painful. Securing capability with colocation providers that already maintain energy and may assist high-density, liquid-cooled {hardware} takes planning measured in quarters. An organization that runs the numbers, concludes it ought to transfer, after which begins wanting is starting a course of moderately than finishing one. The evaluation and the procurement must run in parallel.
The counterweight
Two issues reduce towards every part above, and a bit that ignored them wouldn’t be price studying.
The primary is that inference prices are collapsing. Equal mannequin efficiency has grow to be dramatically cheaper yr on yr, and each month that continues weakens the case for purchasing {hardware} to flee an API invoice which may be considerably smaller by the point the equipment is racked. ICONIQ’s personal margin pattern, from 41% to 52% in two years, is partly this impact. If that curve holds, a few of in the present day’s margin downside solves itself with none infrastructure determination in any respect.
The second is that the majority AI firms shouldn’t be occupied with this. If you’re pre-product-market-fit, iterating weekly, or working variable workloads, the pliability premium is cash effectively spent and infrastructure possession is a distraction from the factor that really kills startups, which isn’t constructing one thing individuals need.
What has modified is narrower than the loudest model of the argument, and extra sturdy. Compute is now a everlasting, rising, structural line in an AI firm’s value of income moderately than a short lived inefficiency to be optimised later. That makes it a board-level query about capital moderately than a procurement element. In a yr when AI has absorbed three-quarters of British enterprise funding, the businesses that perceive their very own compute economics exactly, and the traders who ask about them early, are working with a significantly higher image than those nonetheless making use of a 2021 software program mannequin to a enterprise that now not resembles one.
