Terminal-Bench
Terminal-Bench is an open benchmark project for evaluating AI agents on realistic tasks in terminal and software environments. It publishes tasks, leaderboards, tooling and contributor information. This is a research benchmark resource, not a conventional data annotation employer or a managed labeling service.
Quick facts
- Company
- Terminal-Bench
- Service type
- Data annotation / AI training data
- Specialties
- text, software eng. (code / dev tasks)
- Hiring status
- Application link not verified
- Website
- Website
- Application guide source checked
- Not yet verified
Application process overview
Research and benchmark profile. Visit the official site for its publications, tools and any separately advertised opportunities.
Key findings
Terminal-Bench is included as a benchmark resource. Its contributors page is not a job application or evidence of paid work.