Nexdata offers Multilingual Native & Accented English Speech Data collected from 40 countries, covering a variety of pronunciation habits and topics such as generic, interactive, in-car, and home speech. The dataset includes 40,000 hours of audio data and has 10 years of historical coverage.