KVoiceBench - KIFEval: leaderboard
Metric: Instruction-following accuracy (%; mean of prompt-level and instruction-level accuracy on instructions transferred from IFEval; Korean spoken questions synthesized with TTS; single run at temperature 0.7). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | MiniCPM-o-4.5 | 34.6 |
Interactive version: theaggregate.ai/benchmark?slug=kvoicebench-kifeval · How It Works · Data refreshed daily, snapshot 2026-09-26.