Real software work

See how developers actually build.

Run realistic hiring challenges with inspectable technical evidence, or start a structured software build of your own.

Start Hiring For Developers

Alex Morgan Technical Evaluation showing a 3.60 out of 4 score, complete evidence coverage, and inspectable criteria

Choose your path

Built around real software work.

For hiring teams

Evaluate candidates through realistic software challenges, development evidence, and inspectable Technical Evaluations.

Explore Organizations

For developers

Turn an idea into a structured Solo Marathon with AI planning, milestones, and a completed result.

Explore Developers

Product example

Same challenge. Different evidence. Different outcome.

These example runs show why the same challenge should not produce the same certainty. DevMarathon evaluates the evidence available from each run instead of forcing every candidate into a score.

Alex Morgan Run Signals showing nine pushes, nine commits, one active branch, and a provided submission — **3.60 / 4**Complete evidence

Jamie Chen Run Signals showing one push, one commit, one active branch, and no provided submission — **Insufficient evidence**No overall score

Activity is review context, not a productivity or seniority score.

Follow the evidence

From development context to an inspectable evaluation.

Review recorded GitHub activity from the run, then inspect criterion judgments, confidence, evidence sufficiency, and evidence-linked reference IDs behind the Technical Evaluation.

01 / Activity

See how the work progressed.

Recorded GitHub activity provides review context alongside the submission.

Alex Morgan GitHub activity showing multiple recorded pushes during the run

02 / Evaluation

Inspect the basis of the result.

Technical Evaluation surfaces criterion judgments, confidence, evidence sufficiency, and evidence-linked reference IDs.

Technical Evaluation criterion showing its score, judgment, confidence, evidence sufficiency, and source reference

Human review

AI supports the technical review. Your team makes the decision.

When candidate-attributable evidence is insufficient, DevMarathon can return no overall score instead of inventing certainty.

For developers

Turn an idea into a structured build.

Start with an idea, use AI planning and a starter structure, then work through the build toward a completed result.

Start Building

01Idea
02AI Plan + Starter
03Build + Milestones
04Completed Result

Hiring pricing

Start hiring free. Pay only for additional AI evaluation.

The hiring workflow is free and every organization includes 3 AI Technical Evaluations. Additional evaluations are $15 each.

Start Hiring View Pricing

FAQ

Questions before you start?

What does DevMarathon review in a Hiring Challenge?

DevMarathon is not a coding quiz. Hiring Challenges are built around repository-backed development work. Technical Evaluation uses controlled evidence selected from the submitted work and does not independently browse the entire candidate repository.

Does AI make the hiring decision, and what happens if evidence is insufficient?

No. AI supports the Technical Evaluation; human reviewers make the final decision. When candidate-attributable evidence is insufficient, DevMarathon can return no overall score rather than invent certainty.

Can hiring teams start free?

Challenges and candidate participation are free. Each organization includes 3 AI Technical Evaluations. Additional evaluations are $15 each.

Who needs GitHub?

Organizers can create and manage Hiring Challenges without connecting GitHub. Candidates participating in Hiring Challenges and developers creating their own Solo Marathons do need GitHub for those workflows.

Can developers use DevMarathon on their own?

Yes. Developers can create Solo Marathons with AI planning, a starter structure, milestones, and a completed result.

Start with real work

Review the work behind the submission.

Start Hiring Start Building