The AI job market in Europe: October 2026 report
Tokn's first monthly report on the AI job market in Europe: 774 live roles at 79 companies, with employers, cities, remote share, pay and skills.
Thomas Couderq, Founder of Tokn
12 min read
In this article⌄
- Key numbers on the AI job market in Europe
- Which AI roles are in demand in Europe?
- Shipping models now employs more people than building them
- The new job titles: forward deployed, applied AI, agents
- Forward deployed engineers
- Agents are in the job descriptions, not yet the titles
- Who is hiring: the top 15 AI employers
- A concentrated market
- AI-native companies post most of the roles
- Where the AI jobs are: cities and countries
- Paris and London, neck and neck
- Why Berlin looks small here
- How many AI jobs in Europe are remote?
- Is the AI job market only for senior people?
- What AI jobs pay in Europe, and how few say so
- The skills employers ask for
- How fresh are these openings?
- AI hiring trends in Europe: what other sources say
- Methodology: how we built this report
- What we will watch in the November edition
- Find the roles behind the numbers
- FAQ
- How many AI jobs are open in Europe right now?
- Which companies are hiring the most AI talent in Europe?
- Which European city has the most AI jobs, Paris or London?
- Are AI jobs in Europe remote?
- What do AI jobs pay in Europe?
- Is AI hiring in Europe growing?
The AI job market in Europe, as seen from company career pages, is concentrated, senior and mostly office-based. On Tokn, as of 12 October 2026, 774 AI jobs are open at 79 companies. Ten employers hold half of them, London and Paris are tied, and about one job in six is fully remote.
This is the first edition of a monthly report, and it sets the baseline. Our data starts in October 2026, so every number below is a snapshot, not a trend. Where we mention growth, it comes from other sources, which we link.
In short
- 774 open AI roles at 79 companies on Tokn. Mistral AI alone posts more than one in ten.
- London (224 jobs) and Paris (221) are level. Together they hold three in four jobs tied to a named city.
- 17% of roles are fully remote. Hybrid is more common (28%), and the big labs hire on-site or hybrid.
- About one job in six is a customer-facing deployment role: forward deployed engineer, applied AI, deployment strategist.
- 69% of roles with a stated level are senior or lead. Only about 12% of postings show a salary.
Key numbers on the AI job market in Europe
The headline figures for October 2026, from Tokn's live board.
| Metric | October 2026 |
|---|---|
| Open AI roles | 774 |
| Companies with at least one opening | 79 (of 86 tracked) |
| Share held by the top 10 employers | 51% |
| Largest employer | Mistral AI, 84 roles (11%) |
| London vs Paris | 224 vs 221 |
| Fully remote | 17% (132 roles) |
| Senior or lead, among roles with a stated level | 69% |
| Median minimum experience asked | 5 years |
| Postings that show a salary | about 12% |
| Median euro salary (midpoint of range, 39 postings) | €100,000 |
You can browse all live AI jobs behind these numbers.
Which AI roles are in demand in Europe?
Machine learning and AI operations are the two largest groups, with a fifth of openings each. Classic data science is the smallest. We sort every role into one of eight categories, explained in our guide to the 8 AI career paths.
| Category | Open roles | Share | Fully remote |
|---|---|---|---|
| Machine Learning | 171 | 22% | 13% |
| AI Ops & Automation | 158 | 20% | 16% |
| Research | 91 | 12% | 22% |
| MLOps & Infra | 91 | 12% | 20% |
| LLM & GenAI | 76 | 10% | 26% |
| AI Product & Design | 76 | 10% | 14% |
| AI Marketing & Content | 62 | 8% | 16% |
| Data Science | 49 | 6% | 10% |
Shipping models now employs more people than building them
Roles that build models (machine learning plus research) make up about a third of openings. Roles that deploy, run and productise them (AI ops, MLOps and AI product) make up 42%. We think this is the clearest signal in the data: AI companies in Europe are hiring to get models into customers' hands.
LLM & GenAI is the most remote-friendly category, at 26% fully remote. For a closer look at that category, browse the LLM and GenAI jobs on Tokn.
The new job titles: forward deployed, applied AI, agents
About 130 roles at 27 companies, one in six on the board, are customer-facing deployment jobs. That group combines forward deployed engineers, applied AI roles, deployment strategists and solutions engineers or architects, counted once each.
| Word in the job title | Roles | Companies |
|---|---|---|
| "Applied AI" | 62 | 14 |
| "AI Engineer" | 54 | 24 |
| "Forward Deployed" | 44 | 13 |
| "Research Engineer" | 44 | 13 |
| "Agent" or "Agentic" | 37 | 19 |
| Solutions or sales engineer, architect | 37 | 16 |
| "Deployment Strategist" | 14 | 3 |
| Post-training or reinforcement learning | 12 | 6 |
Forward deployed engineers
A forward deployed engineer works inside a customer's organisation to make an AI product work on their data. On Tokn, 44 titles at 13 companies use the term, at Anthropic, OpenAI, Mistral AI, Cohere, Databricks and Cursor among others. Mistral AI, ElevenLabs and Pigment also hire "deployment strategists".
In the US, Lightcast data cited by Fortune puts US forward deployed engineer postings up more than 1,000% in January to August 2026 against the same period of 2025, from a small base.
Agents are in the job descriptions, not yet the titles
Only 37 titles mention agents, but the word appears in the full text of 218 postings, at 53 of the 79 hiring companies. The same pattern holds for evaluation work: five titles, but 56 postings ask for evals experience.
| Theme in the full posting | Roles | Companies | Share of all roles |
|---|---|---|---|
| Agents or agentic | 218 | 53 | 29% |
| LLMs | 141 | 42 | 19% |
| RAG or retrieval | 63 | 21 | 8% |
| Inference | 58 | 22 | 8% |
| Evals | 56 | 22 | 7% |
| MCP (Model Context Protocol) | 30 | 21 | 4% |
Who is hiring: the top 15 AI employers
Mistral AI, OpenAI, Anthropic, Lovable and Wayve are the five largest AI employers on Tokn, with a third of all openings between them. Twelve of the top 15 are AI-native companies.
| # | Company | AI-native | Open roles | Share | Fully remote |
|---|---|---|---|---|---|
| 1 | Mistral AI | Yes | 84 | 10.9% | 4 |
| 2 | OpenAI | Yes | 48 | 6.2% | 0 |
| 3 | Anthropic | Yes | 44 | 5.7% | 2 |
| 3 | Lovable | Yes | 44 | 5.7% | 0 |
| 5 | Wayve | Yes | 40 | 5.2% | 0 |
| 6 | Celonis | No | 31 | 4.0% | 0 |
| 7 | ElevenLabs | Yes | 30 | 3.9% | 29 |
| 8 | Legora | Yes | 26 | 3.4% | 0 |
| 9 | Cohere | Yes | 25 | 3.2% | 18 |
| 10 | Synthesia | Yes | 24 | 3.1% | 12 |
| 11 | Artefact | No | 23 | 3.0% | 0 |
| 12 | Doctolib | No | 19 | 2.5% | 0 |
| 13 | H Company | Yes | 17 | 2.2% | 0 |
| 14 | Perplexity | Yes | 16 | 2.1% | 2 |
| 14 | Scale AI | Yes | 16 | 2.1% | 2 |
A concentrated market
The top 10 employers hold 51% of openings and the top 15 hold 63%. The typical hiring company has about five open roles. For a job seeker, a short list of companies decides most of the market, so it pays to know all 86 companies we track.
AI-native companies post most of the roles
About three quarters of openings are at AI-native companies, which post 2.3 times more AI roles per company than the others. This partly reflects our list, which leans towards AI-heavy firms. Other employers, such as Celonis, Doctolib and Datadog, are most present in machine learning and data science.
Where the AI jobs are: cities and countries
London and Paris dominate, with almost identical counts. The next hubs each rest on one or two large employers.
| Location | Open roles | Share | Largest employers |
|---|---|---|---|
| London | 224 | 28.9% | Anthropic, Wayve, Lovable, OpenAI, Scale AI |
| Paris | 221 | 28.6% | Mistral AI, Doctolib, Doctrine, Datadog, Dust |
| Fully remote | 132 | 17.1% | ElevenLabs, Cohere, Synthesia, n8n |
| Elsewhere in Europe | 47 | 6.1% | Brussels, Warsaw, Freiburg and others |
| Stockholm | 43 | 5.6% | Lovable, Legora |
| Munich | 26 | 3.4% | Celonis, Anthropic |
| Madrid | 25 | 3.2% | Celonis |
| Amsterdam | 14 | 1.8% | Together AI, Mistral AI |
| Dublin | 13 | 1.7% | |
| Berlin | 11 | 1.4% | Doctolib, Artefact |
| Barcelona | 8 | 1.0% | |
| Zurich | 7 | 0.9% | Anthropic, Artefact |
Paris and London, neck and neck
London has 224 open roles and Paris 221: together, three in four of the roles tied to a named city. London's top employers are mostly US labs plus Wayve and Lovable; Paris is anchored by Mistral AI and French scale-ups. See the live lists of AI jobs in London and AI jobs in Paris.
Why Berlin looks small here
Berlin has only 11 roles on Tokn. That reflects our company list, which has few Berlin-based AI firms so far, not the German market: Indeed data puts Germany ahead of the UK and France for AI-labelled job titles (see below).
How many AI jobs in Europe are remote?
17% of AI roles on Tokn are fully remote, about one in six. Hybrid is more common, at 28%, and four in ten postings don't state a policy.
| Work policy | Open roles | Share of all | Share of roles that state a policy |
|---|---|---|---|
| Hybrid | 214 | 27.6% | 45.9% |
| Fully remote | 132 | 17.1% | 28.3% |
| On-site | 120 | 15.5% | 25.8% |
| Not stated | 308 | 39.8% |
Remote work depends on the employer more than the role. ElevenLabs (29 of 30 roles), Cohere (18 of 25) and n8n (all 9) hire mostly remotely, as does half of Synthesia. The big labs do not: OpenAI has no fully remote roles on Tokn, Anthropic 2 of 44, Mistral AI 4 of 84.
"Remote" also rarely means "anywhere in Europe": many postings say "Remote, UK" or name one country. Browse the current remote AI jobs.
Is the AI job market only for senior people?
Mostly, yes. Among roles with a stated level, 69% are senior or lead, and the median posting asks for at least five years of experience.
| Level (roles with a stated level) | Share |
|---|---|
| Lead (staff, principal, manager, head) | 40.2% |
| Senior | 28.7% |
| Mid | 17.5% |
| Junior | 13.6% |
About a third of postings don't state a level. Among the roughly 280 that state a minimum experience, 12% ask for two years or less, 53% for three to five years, and 35% for six or more.
There is still a way in: about 70 junior roles and about 40 internships. Companies that are not AI-native are about twice as open to juniors (20% of their roles with a stated level, against 11% at AI-natives). Our guide on how to get a job in AI covers the next steps.
What AI jobs pay in Europe, and how few say so
About one posting in eight (11.7%) shows a salary. Across the 39 postings with an annual euro range, the median midpoint is €100,000.
| Currency | Postings | 25th percentile | Median | 75th percentile |
|---|---|---|---|---|
| EUR | 39 | €80,000 | €100,000 | €140,125 |
| GBP | 12 | £114,375 | £115,000 | £156,750 |
Figures are the midpoint of each posted range; we never convert currencies, and 12 pound postings is a small sample. By level, euro medians are €135,250 for lead (11 postings), €110,000 for senior (9) and €58,775 for mid-level (8): indicative only.
Another 34 postings publish a US dollar band (median midpoint $250,825). These come mostly from US-headquartered companies such as OpenAI, Perplexity and Modal that attach US or multi-location bands, so they are not European pay.
Some employers publish pay openly: Elastic, Black Forest Labs (€100k to €340k), n8n (€89.7k to €201.6k), Dust (€40k to €150k) and Anthropic (€125k to €295k). In pounds, Granola (£100k to £150k) and OpenAI (£170k to £445k) do the same.
For context, PwC's 2026 Global AI Jobs Barometer finds an average wage premium of 62% for jobs requiring AI skills, across 27 countries and territories.
The skills employers ask for
Python is close to universal: it appears in 69% of postings where we could extract skills (565 roles, about three quarters of the board). Cloud, Kubernetes and PyTorch come next.
| # | Skill | Roles | Share of roles with skills |
|---|---|---|---|
| 1 | Python | 389 | 68.8% |
| 2 | AWS | 130 | 23.0% |
| 3 | Kubernetes | 126 | 22.3% |
| 4 | PyTorch | 117 | 20.7% |
| 5 | SQL | 109 | 19.3% |
| 6 | GCP | 101 | 17.9% |
| 7 | TypeScript | 98 | 17.3% |
| 8 | Azure | 89 | 15.8% |
| 9 | RAG | 70 | 12.4% |
| 10 | Rust | 65 | 11.5% |
| 11 | Go | 64 | 11.3% |
Three things stand out. PyTorch is asked for 3.5 times more often than TensorFlow (117 against 33), and even JAX (37) is ahead of TensorFlow. Rust (65) and Go (64) both beat Java (43). And TypeScript in the top 10 shows how many AI jobs are product engineering around a model. (We leave React and Node.js out of this table: our skill matching also catches the verb "react" and "nodes", so their counts are inflated.)
How fresh are these openings?
About 36% of live roles were posted or reposted in the last 30 days, and 35% have been open for more than 90 days. Ages use the publication date in each company's applicant tracking system, not when Tokn found the role.
| Posted | Share of open roles |
|---|---|
| Last 7 days | 10% |
| 8 to 30 days | 26% |
| 31 to 90 days | 28% |
| 91 to 180 days | 17% |
| More than 180 days | 18% |
Wayve, Mistral AI, OpenAI, Legora and ElevenLabs posted the most roles in the last 30 days; all of Wayve's openings are less than a month old. Roles open for months are often evergreen or hard to fill.
AI hiring trends in Europe: what other sources say
Tokn can't show growth yet, but others measure it. Their methods differ, so don't compare them directly.
- AI job titles. In Indeed data reported by Euronews, Germany had 288 AI-labelled job titles in Q1 2026, 4.2% of all its job titles, up from 72 in Q1 2022. The UK (160) and France (138) follow. These are shares of job titles, not of postings.
- AI in postings. Indeed Hiring Lab's July 2026 European chartbook finds the share of postings mentioning AI still climbing, led by Ireland (15%), Spain (13%) and the UK (9%), while overall hiring demand weakens.
- AI engineering openings. PredictLeads counts AI engineering openings in Europe up 57.1% between July 2025 and July 2026, from 4,221 to 6,633. One vendor dataset, one comparable month.
- Fastest-growing jobs. LinkedIn's Jobs on the Rise 2026, via Euronews, ranks AI engineer first in the UK, France, Spain and Italy. In Germany, Head of AI is first.
- Employer expectations. In the Linux Foundation's 2026 State of Tech Talent Europe, European organisations expect AI to have a net hiring effect of +27% in 2026. That is a survey of plans, not observed hiring.
Methodology: how we built this report
Tokn tracks about 86 companies hiring for AI in Europe: AI labs, AI-native scale-ups and tech companies with real AI teams. We read their career pages through the public applicant tracking system (ATS) boards behind them, such as Ashby, and refresh every board daily. This report uses the snapshot of active roles taken on 12 October 2026.
Each role is assigned to one of eight categories. Work policy, level, experience and skills come from the postings themselves; title and keyword counts use word matching on titles and full posting text.
What these figures can't tell you:
- The whole European market. Our list is curated and leans towards France and the UK. Germany in particular is under-covered.
- Trends. Our data starts in October 2026, so this edition is a baseline. Month-on-month change starts in November.
- Company size or funding stage. We don't have reliable data on either yet.
- Gaps in postings. About 40% don't state a work policy, a third don't state a level, and only about 12% show pay. Publication dates can reset when a role is reposted.
What we will watch in the November edition
Next month we will report new roles, closed roles and net change against this baseline, by company, city and category. We will also watch:
- Pay transparency. The EU Pay Transparency Directive had a transposition deadline of 7 June 2026. We expect the 12% disclosure rate to rise as countries apply it.
- Deployment roles. Whether forward deployed titles spread beyond the large labs.
- Coverage. New companies on the board, starting with Germany.
Find the roles behind the numbers
The roles behind these figures are live on Tokn: start with AI jobs in London, Paris or remote. The fastest way to act on this report is to create a job alert: pick a category and a city, and we will email you new roles as they appear.
FAQ
How many AI jobs are open in Europe right now?
On Tokn, as of 12 October 2026, 774 AI roles are open at 79 of the 86 companies we track. That is a curated board, not the whole market: PredictLeads, using a broader dataset, counted 6,633 AI engineering openings in Europe in July 2026.
Which companies are hiring the most AI talent in Europe?
On Tokn, Mistral AI leads with 84 open roles, followed by OpenAI (48), Anthropic (44), Lovable (44) and Wayve (40). These five hold a third of all openings, and the top 10 hold half.
Which European city has the most AI jobs, Paris or London?
They are tied. As of 12 October 2026, London has 224 open AI roles on Tokn and Paris 221. Anthropic, Wayve and Lovable lead in London; Mistral AI leads in Paris.
Are AI jobs in Europe remote?
About one in six AI roles on Tokn is fully remote (17%), and 28% are hybrid. Remote roles cluster at a few employers such as ElevenLabs, Cohere and n8n, and many are limited to one country.
What do AI jobs pay in Europe?
Among the 39 postings on Tokn with an annual euro range, the median midpoint is €100,000. Across 12 pound postings it is £115,000. Only about 12% of postings show pay, so treat these as indicative.
Is AI hiring in Europe growing?
For AI roles, external data says yes: PredictLeads counts AI engineering openings up 57% year on year, and Indeed finds a rising share of postings mentioning AI, even as overall hiring weakens. Tokn will measure monthly change from November 2026.
