SpeechIQ: Speech-Agentic Intelligence Quotient Across Cognitive Levels in Voice Understanding by Large Language Models

We introduce Speech Intelligence Quotient (SpeechIQ) as a new form of human cognitioninspired evaluation pipeline for voice understanding large language models (LLM Voice ), designed to assess their voice understanding ability. Moving beyond popular voice understanding metrics such as word error rate (WER), SpeechIQ examines LLM Voice across three cognitive levels motivated by Bloom's Taxonomy: (1) Remembering (i.e., WER for verbatim accuracy); (2) Understanding (i.e., similarity of LLM's interpretations); and (3) Application (i.e., QA accuracy for simulating downstream tasks). We demonstrate that SpeechIQ not only quantifies voice understanding abilities but also provides unified comparisons between cascaded methods (e.g., ASR-LLM) and end-toend models, identifies annotation errors in existing benchmarks, and detects hallucinations in LLM Voice . Our framework represents a first-ofits-kind intelligence examination that bridges cognitive principles with voice-oriented benchmarks, while exposing overlooked challenges in multi-modal training. Our Speech-IQ leaderboard is hosted at huggingface.co/spaces/ nvidia/Speech-IQ-leaderboard.