London is 26.3% of all UK tech hiring and the only part of it where AI demand is reshaping budgets. ON3 Works recruits engineers for London teams with an AI-screened, qualified shortlist in 24 to 48 hours and a senior engineer interviewing every finalist on the Atelier tier. Fees are 7% (Agentic) or 20% (Atelier), no deposit, billed in GBP, against the 18% to 22% London agencies usually charge. And because ON3 recruits in Spain from the same office, a London role can be filled in Madrid or Barcelona on real salary data when the London band doesn't work.
A quarter of UK tech hiring, and most of the AI money
APSCo and CV-Library counted 126,861 unique UK technology vacancies in Q2 2026, up 4.3% on the year, of which 33,312 were in London (26.3%) [APSCo / CV-Library, Jul 2026]. Volume is back, but the average advertised salary across UK tech roles is £71,261 and broadly flat: the money is no longer rising with the market, it is concentrating in a few skill sets.
The clearest of those is AI. AI-related roles are now 4.6% of all UK tech vacancies, up from 3.4% a year earlier, and AI Engineer postings grew 74.4% year on year in permanent hiring (96.3% in temporary roles). Data Scientist salaries rose 22.2%, the strongest increase of any role, and 60% of UK vacancies asking for AI expertise sit in London and the South East [IT Job Board, Jul 2026]. If you are hiring ML or data in London you are bidding in the one segment where pay is still moving up.
Two structural shifts matter for how you run the search. Permanent roles are 88% of tech vacancies and contract hiring fell 6.1% quarter on quarter, so the contractor bench that London teams used to lean on in a pinch is thinner. And employer-direct postings grew 19.8% while recruitment agency volumes fell: companies are trying to hire without agencies, which is exactly why the agencies that remain have to justify their fee with something a careers page can't do, in ON3's case a senior engineer vetting each finalist.
The UK average time to hire is about 4.9 weeks (4.7 in IT), and niche London roles in cloud, security and data run longer still [StandOut CV; NatWest Mentor, 2026]. For the wider UK picture see our UK technical recruiting coverage; if you are weighing a search partner against building the capability in-house, we cost that out in agency versus in-house hiring.
Five London signals and what to do about each
Everything below comes from the APSCo / CV-Library Q2 2026 technology report (Lightcast data, 1.4M+ tech candidates on CV-Library) unless marked otherwise. We have added the practical consequence for a London hiring manager.
| Signal | Figure | What it means for your search |
|---|---|---|
| AI roles share of vacancies | 4.6% (from 3.4%) | ML and data candidates now hold several offers. Decide your band before the first call, and interview inside a week. |
| AI Engineer average advertised salary | £78,821 | Highest of any role and 11% above the all-tech average. Budget for the premium or hire a strong backend engineer and train into it. |
| Adverts that publish a salary | 34% (from 32.1%) | Publishing a real band still stands out in London and cuts the mismatched applications that slow the first stage. |
| Permanent share of vacancies | 88%; contract −6.1% QoQ | The contractor fallback is thinner. If a role is urgent, run the permanent search at full speed rather than waiting for a day-rate bench. |
| Employer-direct postings | +19.8% YoY; agency volumes down | Your careers page is competing with everyone else's. Use an agency only where it adds a vet or a network you don't have. |
Figures: APSCo UK / CV-Library Technology Sector Overview, Q2 2026 data, published 16 July 2026; AI vacancy geography and AI pay premium via IT Job Board's Q3 2026 tracker (ONS, PwC, APSCo, Lightcast). Consequences are ON3's reading, not the report's.
One more number worth knowing: the AI skills wage premium in the UK is put at 34.2%, up from 11% a year earlier [IT Job Board, Jul 2026]. That is the gap you are paying for when a job title says "AI" and the work is mostly backend engineering around a model API. For many London teams the better buy is a senior backend engineer with production LLM experience, interviewed by someone who can tell the difference.
What to budget, by role and seniority
Indicative annual base salary for London from ON3's profile index. Use it to pitch an offer that actually lands; London candidates compare offers in days, not weeks.
| Role | Junior | Mid | Senior |
|---|---|---|---|
| Backend | £31k–48k | £64k–101k | £79k–121k |
| Full-stack | £32k–45k | £45k–75k | £70k–100k |
| ML / AI | £38k–54k | £52k–95k | £86k–142k |
| DevOps / platform | £26k–40k | £50k–80k | £69k–102k |
Source: ON3 Works index of 1.7M+ UK and Spanish technical profiles, refreshed daily (2026). Base salary only. Cross-checks: specialist senior ML engineers in London advertised at £110k to £160k, with generative-AI specialists up to 25% above standard software rates (IT Job Board, Jul 2026). ML and payments-infrastructure roles trend to the top of each band.
What a London agency costs on a £85,000 offer
London tech and IT agencies typically charge 18% to 22% of first-year salary on contingency and 25% to 33% for a retained search. On a £85,000 senior backend offer, the middle of the London band, that is £15,300 to £18,700 on contingency and £21,250 to £28,050 retained, usually with a deposit up front.
ON3's Agentic tier at 7% is £5,950 on the same offer; the Atelier tier at 20%, with a senior engineer interviewing every finalist, is £17,000. Both are billed in GBP with no deposit and a free replacement inside 90 days. The full breakdown of how London fees are built, and what the clauses in a terms-of-business actually mean, is in our guide to recruitment agency fees UK; the recruitment fee calculator runs the same maths on any salary, with VAT and the rebate schedule.
When the London band doesn't work: the same role from Madrid or Barcelona
ON3 recruits in London and Spain from one office, so we see both price lists. The world's largest technology companies already use the second one: of the 238 Spain adverts we coded from the 250 most valuable tech companies on 14 October 2026, 54 from 25 employers are fully remote, 78% of those at senior level or above, and 36 are open across several countries at once. NVIDIA, Akamai, Cisco, Palo Alto Networks and Nebius lead the remote list. The full study is in what big tech is hiring in Spain.
What that costs, senior level, from the ON3 salary calculator (fixed gross pay, before employer contributions; Madrid and Barcelona are within 1% of each other):
| Senior role | London (ON3 index) | Spanish scale-up | International hub in Spain |
|---|---|---|---|
| Backend | £79k–121k | €65.5k–84k | €79.5k–102k |
| Full-stack / frontend | £70k–100k | €61k–78k | €73.5k–94.5k |
| ML / AI | £86k–142k | €76.5k–104.5k | €92.5k–126.5k |
| DevOps / platform | £69k–102k | €76.5k–103k | €92.5k–124.5k |
Convert at the day's rate and read it honestly: for backend and front-end work a senior engineer in Spain at an international hub costs around the bottom of the London band, and at a scale-up well below it. ML narrows the gap, because Spanish ML pay has risen with the same AI demand. DevOps and platform is the exception where the arbitrage mostly disappears; Spain has too few senior platform engineers for a London budget to buy a discount. A London company hiring remotely in Spain pays toward the top of the hub column: candidates benchmark against the big tech adverts above, not against local consultancies.
The practical route is an employer of record or a Spanish entity, English-first teams (234 of the 238 big tech adverts are written in English, and only 2 of the 54 remote ones require Spanish), and a one-hour time difference. ON3 runs the search the same way in both markets; the Spanish side is described on our Spain salary expectations study and the salary calculator (in Spanish).
Two tiers, one engineer-led standard
A precise note on the 24 to 48 hours: that shortlist is AI-screened and qualified, not yet engineer-interviewed. In Atelier, the senior-engineer interview follows before any candidate reaches you. No deposit, billed in GBP, free replacement inside 90 days. Both tiers and worked examples are on the pricing page; the difference between this and a conventional London agency is set out in AI recruiting versus a traditional agency.
From payments backend to deep tech
Strong in software across the stack, growing fast in ML, cybersecurity and deep tech (chips, HPC). Each role type is broken down in the profiles we cover directory. The most common London searches:
- Backend (Python, Go, Java)
- ML / AI engineering
- Data engineering
- DevOps / platform
- Full-stack
- Frontend / React
- Payments infrastructure
- Cybersecurity
- Engineering leadership
Who ON3 hires for in London
Two kinds of London buyer come to ON3. Teams worn down by agencies charging a fifth of salary for a stack of CVs want a real technical vet: finalists interviewed by an engineer who has built in that stack. And fast-moving teams that never used agencies on cost try the Agentic tier at 7%. In London the first group is bigger, which is why the engineer-led interview is the part of the service we lead with.
One example from ON3's venture-capital channel: a YC-2023 company that raised a Series B in 2026 used ON3 to hire 12 technical staff, fast. CGI, which uses ON3 for 50+ hires a year, operates in the UK as well as Spain.
“ON3 Works has become my favourite external recruiting partner. They're fast, efficient, and bring us candidates already vetted, qualified, and ready to hire. They've made talent search easy for me and my team.”
— Gabriel Garcia, Vice-President Consulting Services, CGI
London hiring FAQ
How much do tech recruitment agencies charge in London?+
How big is the London tech job market right now?+
What does a senior backend engineer earn in London in 2026?+
Should I publish the salary in a London tech advert?+
Can a London company hire the same engineers in Spain instead?+
How fast can ON3 fill a London engineering role?+
Hiring in London? Tell us who you need
Send us the role and the band. You'll get a qualified shortlist in 24 to 48 hours, and for critical hires, a senior engineer who interviews every finalist. If the London band doesn't work, we'll price the same role in Spain.
Tell us who you need to hire →Sources & notes
UK and London market APSCo UK / CV-Library, Technology Sector Overview (Q2 2026 data, Lightcast; published 16 July 2026): 126,861 UK vacancies (+4.3% YoY), 33,312 in London, average advertised salary £71,261, AI roles 4.6% of vacancies (from 3.4%), AI Engineer postings +74.4% permanent / +96.3% temporary, AI vacancies +41.8%, AI Engineer average advertised salary £78,821, Data Scientist pay +22.2%, permanent 88% of vacancies, contract −6.1% QoQ, employer-direct +19.8%, 34% of adverts disclose salary. IT Job Board, UK Tech Salary Tracker Q3 2026 (29 July 2026): 60% of AI-expert vacancies in London and the South East; senior ML engineer London £110k–160k; AI wage premium 34.2% (from 11%). Time to hire: StandOut CV (~4.9 weeks overall, 4.7 in IT; survey of 497 recruiters); NatWest Mentor (Apr 2026; niche tech roles stretch longer).
Salary bands London: ON3 Works index of 1.7M+ UK and Spanish technical profiles (2026), cross-checked against IT Jobs Watch and the IT Job Board tracker. Spain: ON3 salary calculator (synthesis of the 2026 Hays, Michael Page, Robert Walters, Manfred, Randstad and Tecnoempleo guides), senior level, Madrid, 'Spanish scale-up' and 'international hub' employer tiers, fixed gross pay. Remote big tech figures: ON3's coding of every Spain advert from the 250 most valuable tech companies, observed 14 October 2026.
ON3 figures and note 7%/20%, no deposit; 24 to 48h shortlist; <30-day median hire; 90.7% retention; 90-day replacement; YC/Series-B and CGI references: supplied by ON3 Works. The 24 to 48h shortlist is AI-screened-qualified, not yet engineer-interviewed; the engineer vet follows in Atelier. Fee examples use 18–22% contingency and 25–33% retained as typical London rates; your quote may differ. Salary figures are indicative and should be confirmed for a specific role.