OpenAI has combined several engineering, product and research teams to rebuild its audio models as it prepares an audio-first personal device expected to launch in about a year, according to reporting by The Information. The company’s new audio model, slated for early 2026, is said to sound more natural,