VoiceBench

VoiceBench is a benchmark that evaluates the performance of audio-language models on various tasks related to auditory perception and instruction-following.

Papers using VoiceBench (7)

DeSTA2.5-Audio: Toward General-Purpose Large Audio Language Model with Self-Generated Cross-Modal Alignment2025 · 46 citesKVoiceBench, KOpenAudioBench, and KMMAU: Agent-Driven Korean Speech Benchmarks for Evaluating SpeechLMs2026AURA: Agent for Understanding, Reasoning, and Automated Tool Use in Voice-Driven Tasks2025 · 1 citesDuplexCascade: Full-Duplex Speech-to-Speech Dialogue with VAD-Free Cascaded ASR-LLM-TTS Pipeline and Micro-Turn Optimization2026DIFFA: Large Language Diffusion Models Can Listen and Understand2025DrVoice: Parallel Speech-Text Voice Conversation Model via Dual-Resolution Speech Representations2025VoiceBench: Benchmarking LLM-Based Voice Assistants2024