[KongchangAI]
ModelWhisper Large V3 Turbo / Whisper Turbo

Whisper Large V3 Turbo

A speech recognition model released by OpenAI as part of the Whisper series. It supports multilingual transcription and is designed for fast operation at low cost.

Timeline (last 90 days)

Oct 22

A certain speech recognition model is only 164MB in size and transcribes one hour of audio in about 20 seconds on an ordinary thin-and-light laptop, with a word error rate of 5.21, outperforming mainstream models such as Whisper Large V3 Turbo; it currently supports only English.

Unverified50%
Oct 20

The project uses the OpenAI Whisper Large V3 Turbo model for speech transcription, and in measured tests transcribing one sentence costs less than one-thousandth of a cent.

Unverified50%
Oct 20

Whisper Large V3 Turbo improves inference speed by reducing the decoder layers from 32 to 4 while maintaining recognition accuracy close to the original.

Unverified50%
Oct 18

The project uses the Whisper Large V3 Turbo model, sending Base64-encoded audio data to the /audio/transcriptions endpoint for transcription.

Unverified50%
Oct 6

Parakeet Redux can run on CPU with a disk footprint of 178 MB, compared to the 1.6 GB of Whisper Large V3 Turbo (Whisper CPP version).

Unverified50%

All Facts (5)

Source Articles