Anthropic careers in Europe: roles, offices and how to apply
Anthropic hires in London first, then Munich, Dublin, Zurich and Paris. Which roles, what postings ask and pay, office rules and how it interviews.
Thomas Couderq, Founder of Tokn
9 min read

In this article⌄
- Anthropic in 30 seconds
- What Anthropic is hiring for in Europe
- Applied AI architects and engineers
- Forward deployed engineers
- Infrastructure and security engineering
- Research engineering and Fellows
- Where the jobs are: London, Dublin, Munich, Paris, Zurich and Milan
- Anthropic London jobs
- Anthropic Dublin jobs
- Munich, Paris, Milan and Zurich
- Office policy
- Seniority and experience
- What Anthropic's job postings ask for
- The Anthropic hiring process
- Using Claude to apply: Anthropic's rules
- Anthropic salary and benefits in Europe
- How to stand out
- Find Anthropic jobs on Tokn
- FAQ
- Does Anthropic hire in Europe?
- Can I work for Anthropic remotely from Europe?
- Does Anthropic sponsor visas?
- How much does Anthropic pay in London?
- Does Anthropic offer internships?
- Can I use Claude in my Anthropic application?
Anthropic careers in Europe are concentrated in London. As of 14 October 2026, we track 50 open AI and technical roles at Anthropic on Tokn. Two in five are customer-facing roles that deploy Claude at companies, one in three is infrastructure or security engineering, and eight are research engineering.
Anthropic and OpenAI each list about 50 AI and technical roles on Tokn, behind Mistral AI, in our ranking of the AI companies hiring the most in Europe. Compared with OpenAI's European hiring, it has fewer customer roles and more research. Tokn is independent of Anthropic; this guide follows our OpenAI and Mistral AI guides.
In short
- 50 Anthropic roles on Tokn as of 14 October 2026, 13 posted in the last 30 days. Its own board had 80 European postings, including sales and finance.
- 20 of 50 are customer-facing (applied AI and forward deployed), 16 are engineering, 8 are research engineering.
- London 33, Munich 5, Dublin 4, Zurich 4, Paris 3, Milan 1.
- Every posting asks for at least 25% of your time in an office. No role is fully remote.
- Pay ranges on most London, Paris and Dublin postings, none in Munich, Zurich or Milan. Customer roles outside London need the local language.
Anthropic in 30 seconds
Anthropic is the San Francisco company behind Claude. Europe is about one in eight postings on its board.
| Founded | 2021 (first announcement, May 2021) |
| Structure | A public benefit corporation, headquartered in San Francisco |
| Products | Claude apps, the Claude Developer Platform (API), Claude Code |
| Funding | $65bn Series H at a $965bn post-money valuation, May 2026; run-rate revenue above $47bn, per Anthropic's announcement |
| Size | No official global figure. More than 200 people in London (press, 2026) |
| European offices | London (2023), Dublin (2024), Zurich (2025), Paris and Munich (announced November 2025), Milan (announced May 2026) |
| Job board | 652 postings worldwide on 14 October: 78% in the US, 12% (80) in Europe (Anthropic jobs) |
What Anthropic is hiring for in Europe
Mostly engineers who deploy Claude at customers or run the infrastructure that serves it. The table sorts the 50 roles on Tokn by title.
Tokn's Anthropic page lists AI and technical roles, not the whole board. Anthropic's board had 80 European postings on 14 October; most of the 30 we don't list are sales and customer success (18), the rest finance, recruiting, legal and compute roles.
| Family | Roles | Examples |
|---|---|---|
| Applied AI architects and engineers | 14 | Applied AI Architect (Industries, Partnerships, Startups), Applied AI Engineer (Enterprise, Startups) |
| Infrastructure engineering | 10 | Staff Software Engineer: Inference, AI Reliability, Kubernetes Platform, Continuous Integration |
| Research engineering | 8 | Pretraining, Reinforcement Learning, Post-Training, Cybersecurity RL |
| Forward deployed engineering and delivery | 6 | Forward Deployed Engineer, Technical Deployment Lead |
| Security and safeguards | 4 | Staff Software Security Engineer, Detection & Response, Safeguards Infrastructure |
| Other | 8 | Fellows tracks (2), Regional Research Economist, Connectivity engineering (2), IT and AV engineers (3) |
Applied AI architects and engineers
The largest family: 14 roles in London, Munich, Paris and Milan. Architects work pre-sales, aiming to become "a trusted technical advisor" to customers, and need 5+ years as a solutions architect, sales engineer or technical account manager.
Applied AI Engineers are more hands-on: "pair programming, architecture reviews, and code contributions" in the Startups role. They asked for 4+ years (Startups) or 6+ years (Enterprise), and welcomed former technical founders.
LLM roles in London, at Anthropic and elsewhere:
Forward deployed engineers
Six roles: FDEs in London, Paris and Munich, plus three delivery leads and managers in London. The FDE "embeds directly with our most strategic customers" and delivers "MCP servers, sub-agents, and agent skills". Expect 25–50% travel. London asked for 4+ years, Paris and Munich for 8+. See what a forward deployed engineer does.
Anthropic's Claude Frontier Academy (October 2026) will train 10,000 FDEs at customers and partners by the end of 2027. That is a training programme, not jobs at Anthropic.
Infrastructure and security engineering
16 roles, almost all Staff level in London and Dublin: inference, reliability, Kubernetes, CI, cluster infrastructure, security and two Connectivity roles. A Dublin inference posting said LLM inference experience was "highly encouraged but not strictly necessary". Rust appears in 11 of the 50 roles, and 11 mention on-call.
Research engineering and Fellows
Eight research engineer roles: six in London (pretraining, reinforcement learning, alignment) and two in Zurich (post-training, cybersecurity RL). With two Fellows tracks and an economist, research is 11 of 50, against 5 at OpenAI. See research scientist vs research engineer.
The alignment posting said the team's hub is in San Francisco, "so you may not hear back on your application to the London team unless we see an unusually strong fit."
Where the jobs are: London, Dublin, Munich, Paris, Zurich and Milan
London leads by far. Like OpenAI, Anthropic posts one city per posting.
| City | On Tokn | Anthropic board, all functions | Office |
|---|---|---|---|
| London | 33 | 49 | Since 2023 |
| Munich | 5 | 12 | Announced November 2025 |
| Dublin | 4 | 11 | Since 2024 |
| Zurich | 4 | 4 | Since early 2025 |
| Paris | 3 | 5 | Announced November 2025 |
| Milan | 1 | 1 | Announced May 2026 |
Anthropic London jobs
London was Anthropic's first European office, opened in 2023, Euronews reported. Two of the 33 London roles are really US-first: one prefers candidates in New York, Seattle or San Francisco, the other lists US cities and a dollar band. Anthropic has 33 of the 235 AI jobs in London on Tokn.
A new London office will hold about 800 staff, against more than 200 today, CoStar reported in April 2026; Anthropic hasn't announced it itself.
Anthropic Dublin jobs
Dublin opened in 2024, per Euronews, and is home to Anthropic Ireland Limited, its EU entity. On Tokn, Dublin roles are inference and reliability engineering plus IT support. In March 2026 Anthropic announced 200 new Irish jobs, TechCentral reported. See all AI jobs in Dublin.
Munich, Paris, Milan and Zurich
- Munich: announced with Paris in November 2025. Five applied AI and FDE roles on Tokn; its other postings are sales. Four of the five require native, C1 or fluent German. See AI jobs in Munich.
- Paris: three applied AI and FDE roles, posted in French and English. Two require native or fluent French.
- Milan: Anthropic's sixth European office, announced in May 2026. One architect role, requiring native or C1 Italian.
- Zurich: a small research-focused office led by Neil Houlsby. Two research engineers, a security engineer and an IT engineer: 4 of the 11 AI jobs in Zurich on Tokn.
Office policy
Every European posting says: "we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices." One London research role asks for five days a week. OpenAI's postings ask for three days a week.
Anthropic's careers page adds that "Most staff are in the Bay Area." No role on Tokn is fully remote; if you need that, see remote AI jobs in Europe. Postings say Anthropic sponsors visas, but not "for every role and every candidate".
Seniority and experience
Anthropic hires experienced people: 19 of the 50 titles are Staff, Senior, Manager or Lead, and none is junior. Stated minimums include:
- 4+ years: Applied AI Engineer (Startups), FDE in London.
- 5–6+ years: Applied AI Architect; Applied AI Engineer (Enterprise).
- 8+ years: FDE in Paris and Munich; Manager, Technical Deployment.
- 10–12+ years: most Staff infrastructure roles.
Research postings state no year count. "We don't currently offer internships," the careers page says. The early route is the Anthropic Fellows Program: a 4-month fellowship, not a permanent job, with a stipend of 2,310 GBP a week, compute funding and a London option. The next cohort is expected in January 2027.
What Anthropic's job postings ask for
LLM work, customers and evals. In the full text of the 50 postings:
- LLM work: 34 mention LLMs, 25 evals, 20 agents. None mention RAG.
- Customers: 26 mention customers or clients, 19 travel, 13 pre-sales.
- Code: 33 mention Python; Rust (11) and Kubernetes (13) mark the infrastructure roles.
- Safety: 22 mention it. Only 2 mention a PhD.
- Local language: 7 roles require German, French or Italian, often at native level. London and Dublin roles ask for English only. That matches OpenAI, unlike Mistral.
The Anthropic hiring process
Anthropic's careers page describes the process, without durations or a round count:
- Apply on the board. Put independent research, blog posts or open-source work at the top of your CV.
- Interviews on Google Meet: live coding in Colab or CodeSignal for technical roles (two research postings said all interviews are in Python), conversations for the others.
- After a no: Anthropic gives no feedback on CVs or interviews, and you can re-apply after 12 months.
To practise, use our ML and AI interview questions.
Using Claude to apply: Anthropic's rules
Anthropic's guidance on candidates' AI use (updated 10 July 2025) sets the defaults:
- Write the first draft of your application yourself; you may then use Claude to refine it.
- Use Claude to prepare for interviews.
- No AI in take-home assessments or live interviews unless Anthropic says otherwise.
- Don't use it to invent experience, and be transparent about your use.
OpenAI's guide, by contrast, says expectations "vary by interview".
Anthropic salary and benefits in Europe
Anthropic shows an "annual compensation range" on most London, Paris and Dublin postings: 35 of the 50 on Tokn. Munich, Zurich and Milan roles show none.
| London posting (advertised in October) | Annual salary range |
|---|---|
| Applied AI Architect | £150k–£190k |
| Applied AI Engineer, Forward Deployed Engineer | £225k–£255k |
| Deployment leads and managers | £165k–£325k |
| Staff Security Engineer; Engineering Manager, Connectivity | £255k–£325k |
| Staff infrastructure (Inference, Reliability, CI) | £325k–£390k, up to £485k for some |
| Research engineer (Pretraining, RL) | £260k–£630k |
The research engineer ranges, which go up to £630k–£640k, are at the top of the European market. Two London-listed roles show only dollar bands: US pay, not London pay.
| Euro posting (advertised in October) | Annual salary range |
|---|---|
| Paris: Applied AI Architect, Partnerships | €190k–€220k |
| Paris: Applied AI Engineer, Enterprise | €220k–€235k |
| Dublin: Senior SE, Inference; Staff SE, AI Reliability | €235k–€295k |
| Dublin: Staff Software Engineer, Inference | €295k–€355k |
| Dublin: IT Support Engineer | €125k–€165k |
Staff Software Engineer, Inference was £325k–£390k in London and €295k–€355k in Dublin. OpenAI showed ranges on three London postings only.
The careers page lists 22 weeks of paid parental leave, an education stipend and relocation support; benefits vary by country. For other employers, see what AI roles pay in Europe.
How to stand out
Based on the postings and Anthropic's pages, we'd focus on:
- Lead with public work. The careers page asks for research, writing or open source at the top of your CV.
- Show production LLM work with evals. 25 of 50 postings mention evals.
- Bring customer-facing or founder experience for applied AI and FDE roles.
- Check the language line on Munich, Paris and Milan roles.
- Practise live Python coding without AI. That is the default in interviews.
- Read the location line. Some London-listed teams prefer US-based candidates.
- Consider the Fellows Program if you're early in your career.
Find Anthropic jobs on Tokn
Anthropic posted 13 roles in the last 30 days; 23 of its 50 were first published over 90 days ago. See every current role on the Anthropic jobs page, browse AI jobs in London, and set a job alert to hear about new ones first.
FAQ
Does Anthropic hire in Europe?
Yes. As of 14 October 2026, Tokn lists 50 Anthropic AI and technical roles in Europe. Its own board had 80 European postings, about 12% of 652 worldwide, including sales and finance roles Tokn doesn't list.
Can I work for Anthropic remotely from Europe?
No. Every European posting asks you to be in an office at least 25% of the time, and some roles ask for more: one London research role asks for five days a week.
Does Anthropic sponsor visas?
Its postings say it does, but not "for every role and every candidate". After an offer, it says it makes "every reasonable effort" and retains an immigration lawyer.
How much does Anthropic pay in London?
Ranges advertised in October were £225k–£255k for applied AI engineers and FDEs, £325k–£390k (up to £485k) for Staff infrastructure and £260k–£630k for research engineers. Munich, Zurich and Milan postings show none.
Does Anthropic offer internships?
No: "We don't currently offer internships," says the careers page. The Anthropic Fellows Program is a 4-month fellowship with a weekly stipend and a London option, not a permanent job.
Can I use Claude in my Anthropic application?
Partly. Anthropic's guidance lets you use Claude to refine a draft you wrote and to prepare for interviews. It doesn't allow AI in take-home assessments or live interviews unless you are told otherwise.





