xAI

xAI publishes a four-step process on its careers page: application review (CV plus a written "statement of exceptional work"), a short screening interview, technical interviews, and an offer. The firm states "we generally do not use recruiters for assessments" — applications are evaluated by xAI's technical team members. The University of Miami's Toppel Career Center describes the same shape in more detail: a ~15-minute screening call followed by a main round of four technical interviews (coding in a language of choice, two domain-specific sessions, and a "Meet the Team" presentation of past work), with the firm targeting completion within one week.

The exceptional-work statement is a real upfront artifact, not a checkbox, and the absence of a separate behavioural round shifts the assessment onto demonstrable technical execution. A candidate should prepare a focused past-work write-up early and plan for the main loop to land as a single concentrated week rather than a multi-month process.

xAI's careers page states under benefits: "Visa sponsorship — We support international talent with visa sponsorship to join our team." Public LCA records confirm xAI's main US operating entities sponsor H-1B workers at scale: xAI Corp filed 7 LCAs in FY2025, and X.AI filed 39 LCAs plus 47 Form I-129 H-1B petitions, all approved (H1BGrader).

Sponsorship is a firm-wide commitment rather than role-by-role discretion, so a sponsorship-required candidate generally does not need to pre-filter postings — but xAI prioritises in-person work across its US and London offices, so a US-located visa remains the practical path for most engineering roles.

Two facts shape what working at xAI looks like. Compute concentration is the first: xAI's Memphis Colossus supercomputer scaled from 100,000 GPUs at its July 2024 launch to 200,000 GPUs by mid-2025, with HPCwire and NVIDIA's own announcement describing it as the world's largest single AI training cluster — research and infrastructure hires work directly on a system at that scale rather than on shared cloud quota. Second, hiring is run end-to-end by the technical team rather than by recruiters (xAI's careers page is explicit on this), and the firm prioritises in-person work across its Palo Alto, Seattle, Austin, New York, Memphis and London offices, with a stated preference against remote engineering roles.

A candidate weighing xAI against Anthropic, OpenAI or Google DeepMind should treat the in-person requirement and the Memphis data-centre footprint as load-bearing facts: a substantial slice of postings — data centre, infrastructure, model training — exists because of the Colossus build-out, so a geography or remote-only preference will close off most of the open roles.

xAI currently has no open internship or graduate roles listed on Jorb AI. This page updates continuously; check back as new postings rotate.

Applications to xAI go through the official careers site. Each role on this page links directly to the specific application. xAI does not use third-party submission services.