Overview
Coarena by Coasty is a platform that provides real-world evaluations for computer-use agents. It features live tasks and blind human judgment, and it publishes every number along with the rule that produced it. The platform is backed by Y Combinator.
Users can watch two computer-use agents work on their research. The site includes a leaderboard, benchmarks, and a CUA KnowledgeBench for comparing models. It also offers model profiles, a dataset, methodology, a safety benchmark, and a metrics API.
The platform emphasizes transparency, stating that every claim on the site cites the file that keeps it. It also covers saved credentials and agent memory in its terms of use and privacy policy.
Pricing
Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.
Visit WebsitePrices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.
Use Cases
- Teams use Coarena to watch two computer-use agents work on their research.
- Teams use Coarena to perform real-world evaluations of computer-use agents.
- Teams use Coarena to compare models using benchmarks.
- Teams use Coarena to benchmark computer-use agents with CUA KnowledgeBench.
- Teams use Coarena to assess the safety of computer-use agents.
Features
Submit real browser tasks
Compare two frontier AI agents
Watch agents race live
Choose winners through blind judging
Votes update computer use leaderboard
Reviews
No reviews yet. Be the first to share your experience!