This model is over 2 years old. The same company has since released something newer: GPT-5.6-Cyber (2026-08-10). The details below are still accurate for this model — it just is not what we would recommend today.

Whisper large-v3

Name: Whisper large-v3
Author: OpenAI

OpenAI🇺🇸 United States

active

Compare with other models →

Version History

large-v3majorNovember 6, 2023

Whisper large-v3 increases spectrogram resolution from 80 to 128 Mel bins, adds a dedicated Cantonese language token, and was trained on 5 million hours of weakly and pseudo-labeled audio, cutting word error rates 10-20% versus large-v2.

Coverage

model releaseOpenAI

OpenAI Releases Whisper Large-v3, Cutting Speech Recognition Errors 10-20% Across Languages

OpenAI has released Whisper large-v3, an open-weight automatic speech recognition and translation model trained on 5 million hours of audio. The model reduces transcription errors by 10-20% compared to its predecessor and adds native Cantonese support.

August 11, 2026 · 5:36 AM2 min read

Whisper speech-to-text ASR