Last updated 2026-07-22

Terms of Service

Acceptance

By using HumanBench, you agree to these terms. HumanBench provides benchmark methodology and comparative evidence status for informational purposes.

Service

A protocol, score or future ranking is comparative evidence, not proof about one person, document or provider. Rankings remain unavailable until the published qualification boundary is satisfied.

No text-checking service

HumanBench does not accept user drafts or provide individual AI-writing detection. Links to checking or revision products leave this domain and are governed by the destination's terms.

Use

Do not use benchmark information as an automatic accusation, authorship decision or way to circumvent school, workplace or publication policy.

Liability

HumanBench is provided as-is without warranty. You remain responsible for how you interpret and use benchmark information.