Mercor domain application strategy: which track to apply into

Mercor accepts applicants into specific tracks (general code review, agent-trace evaluation, formal reasoning, multimodal, legal, medical, etc.). The track you apply into shapes both your acceptance probability and the rate you can command. Picking deliberately matters. ## Map your depth honestly Write down what you have actually built or worked on for at least 18 months. The track that maps to that work has the highest acceptance probability. A Python backend engineer applying into code-evaluation passes faster than the same person applying into formal mathematics, even if they enjoy maths. ## Pick the scarcest track within your depth Among tracks that fit your background, the one with the thinnest applicant pool pays more. In 2026 the scarcest tracks for software-aligned profiles are formal verification, systems languages (OCaml, Rust kernel work, Solidity), and specialised security review. ## Avoid generic tracks if you have specifics General code review is the most-applied-to track and has the lowest hourly rate. If you have any specifics (a language, a framework, a domain), apply into the matching narrow track instead. ## Multi-track applications Mercor lets you apply into multiple tracks. Submit one narrow specialist track first, then one broader fallback. Acceptance into the narrow track unlocks invitations into the broader one anyway; the reverse is rarely true. ## Re-application If you are rejected from a track, you can usually re-apply after six months. Between attempts, publish work that demonstrates the specific depth you applied with (open-source contributions, a technical blog post, conference talks). Profile depth between attempts is the strongest re-application signal. ## What disqualifies even strong candidates Generic AI-rewritten essays. Mercor screens for human judgment, and LLM-assisted essays read as exactly the inverse. Write your application in your own voice with specific examples; that alone separates you from half the pool.