Universal-2

Speech-to-text model from AssemblyAI that transcribes real-world conversations with improved accuracy for developers building voice products.