TY - RPRT TI - Benchmarking Speaker Embedders in the Sub-Second Regime for Real-Time Voice AI Agents AU - Abhishek Gautam PY - 2026 UR - https://arxiv.org/abs/2611.09516 ID - 2611.09516 ER -