Informal comparison
Bohita Long-Session Voice Agent Evaluation
Compare voice agents over a conversation long enough for per-turn behavior and costs to change.
Explore the original evidence
Explore the original evaluation, its methods, and reported findings.
Open evalCoverage
The publisher’s Turn 40 study follows voice agents across a long session and links its voice-benchmark repository. It asks what short, single-response tests leave out.
Methodology
Compare behavior later in the conversation with the early turns and consider the cost over the whole session. Bohita is a participating vendor. This tests a particular session setup, not general speech recognition or end-of-turn detection.