Groq Launches OpenAI's Whisper Large v3 Turbo Model Endpoint with 216x Speed, 50% Smaller Size

Oct 9, 2024, 05:13 PM

Groq has launched the endpoint for OpenAI's new Whisper Large v3 Turbo Speech-to-Text model. The new 'Turbo' version of Whisper Large v3, released last week by OpenAI, is nearly 50% smaller than the non-Turbo Large v3 model, reducing the parameter count from 1.55 billion to a smaller size. Groq's implementation of the model boasts a 216x real-time speed factor, as benchmarked by ArtificialAnlys. The model is now available for use on Groq's platform, and the company is offering private dedicated instances for interested users. Wednesdays are for Whisper, and users can now play with the model on GroqInc.