Whisper is OpenAI's open-source, general-purpose automatic speech recognition (ASR) model. Released as a Transformer sequence-to-sequence system trained on large-scale weakly supervised audio, it handles multilingual transcription, speech translation into English, and spoken language identification from a single set of weights. The code, model weights and command-line tools are published on GitHub under the MIT license, so anyone can run local transcription without paying per minute.
Whisper is OpenAI's open-source, general-purpose automatic speech recognition (ASR) model. Released as a Transformer sequence-to-sequence system trained on large-scale weakly supervised audio, it handles multilingual transcription, speech translation into English, and spoken language identification from a single set of weights. The code, model weights and command-line tools are published on GitHub under the MIT license, so anyone can run local transcription without paying per minute.
