SOURCE SUMMARY
GPT-6 Astra tops Retell’s voice agent benchmark, while Codex helps it reach a $1.6M annual revenue run rate per employee.
WHY IT MATTERS
This update makes a measurable performance claim, which is more useful than a generic capability statement if the comparison is reproducible. It can influence model or workflow selection only when the baseline, workload, and evaluation setup are comparable to real use.
WHAT TO VERIFY
Check the baseline, sample size, task mix, model and effort settings, token or tool costs, evaluation harness, and whether the reported result comes from controlled testing or a single customer case study.