OpenAI Launches Whisper V3 Turbo with 8x Faster Speed for DevDay

Oct 1, 2024, 06:19 AM

OpenAI has announced the launch of its latest AI models designed to significantly enhance problem-solving capabilities in science, coding, and mathematics. Among these new models is Whisper V3 Turbo, an optimized version of the Whisper Large V3 model. Whisper V3 Turbo offers 8x faster transcription speed compared to Whisper Large, 4x faster than Whisper Medium, and 2x faster than Whisper Small, while maintaining minimal degradation in accuracy. The model features 809 million parameters and full multilingual support. OpenAI released these models in time for DevDay, highlighting their commitment to advancing AI technology. The Whisper V3 Turbo model is open-source and was updated on Github yesterday.