Rev, a leading speech technology company, has introduced new AI models, Reverb ASR and Reverb Diarization, which are setting new standards in speech transcription and diarization. These models, Reverb ASR v1 and v2, are open-weight and open-source, and have been reported to outperform current state-of-the-art (SoTA) models. The Reverb ASR model is built within the WeNet framework, while the Reverb Diarization model is developed within the Pyannote framework. The research team at Rev, including Nishchal Bhandari, Danny Chen, and Miguel Ángel del Río Fernández, has focused on these models to enhance accuracy and efficiency in speech recognition and diarization, marking a significant advancement in AI-driven speech technology.