Backed by Y CombinatorLooking for someone who can build a landing page?
Let them show you. Litmus turns your repos, tickets, and job descriptions into real hiring signal.
Building teams for
WHAT LITMUS IS
Litmus generates async coding assessments from a company's own codebase, tickets, and job descriptions: real tasks, scoped to show how a candidate would actually contribute to your team.
Build an assessment from this ticket
Litmus building now
Success
WHY TEAMS SWITCH
Hirehowyou'vealwayswantedto.
codex · GPT-5.2THE CANDIDATE EXPERIENCE
Candidates complete end-to-end feature builds in their native setup, with all of their AI tools available. Litmus tracks everything from file iteration, to commit history, to AI prompts.
THE REPORTS
Seehowtheybuild,notjustwhattheyshipped.
Every submission runs live sandbox against a deterministic harness for correctness. Configurable LLM graders evaluate qualitative dimensions like process and architecture, then compile results into an instantly shareable report.
WHAT CANDIDATES SAY
The assessment candidates want to take.
“Should become the standard OA method. Well scoped, and a lot of fun!”
Stop losing good candidates to broken pipelines.
FAQs
Litmus fits at whatever stage you need a high-signal read on how someone actually builds. Two common ways teams slot it in: up front, for a trustworthy measure of technical ability before synchronous interviews, or in later rounds, to capture process signals like AI prompt usage and coding behavior alongside raw output. Book a demo and we’ll find your fit.
Yes, and we encourage it. Candidates use the same AI tools they’d have on the job, and Litmus tracks every prompt, and analyzes how they reasoned, prompted, and refined: signal you can’t get from output alone.
Every submission runs against a rigorous deterministic harness, specific to your assessment, in a live sandbox for correctness. Configurable LLM graders then evaluate qualitative dimensions like architecture and process, and every judgment is tagged to the file and line that backs it up.
Yes. Bring your own take-home or coding problem and we’ll host and grade it, or let Litmus generate one from your codebase, tickets, and job descriptions. Either way you get the same evidence-based report.
That’s fine. Connecting your codebase and tickets makes assessments sharper, but none of it is required. Litmus is fully configurable: share as much or as little as you want, and we can build a relevant assessment from public inputs alone, like your job descriptions. Either way, we’ll work with you to shape it to your team’s context.
Yes. Candidates sync to and from Ashby, Greenhouse, and Lever automatically, or you can manage them end to end on Litmus. Nothing about your pipeline has to change.
Litmus tracks file iteration, commit history, and AI prompts across the whole build, and every candidate records a short walkthrough explaining their decisions. That makes an assessment far harder to game than a standard take-home.
We price custom contracts according to your hiring needs and volume. Reach out to us at founders@litmushiring.com to learn more!
Book a demo and we’ll build your first assessment from your own context. Most teams are sending assessments to candidates within a few days.