EVA-Bench - Task Completion: leaderboard
Metric: Task completion (%; share of trials whose final scenario-database state matches the ground truth exactly; 213 scenarios in three enterprise domains, five clean-audio trials each, averaged equally over the airline, healthcare HR and IT service management domains). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 4 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | gpt-realtime-1.5 | 73.9 |
| 2 | Ultravox-Realtime | 47.3 |
| 3 | Gemini 3.1 Flash Live | 47.3 |
| 4 | GPT-Realtime-mini | 34.5 |
Interactive version: theaggregate.ai/benchmark?slug=eva-bench-task-completion · How It Works · Data refreshed daily, snapshot 2026-09-26.