Whisper

Whisper is OpenAI's general-purpose automatic speech recognition (ASR) system, released open-source in September 2022 with full model weights and code under the MIT license. It is a Transformer sequence-to-sequence model trained on 680,000 hours of multilingual, multitask supervised audio, and a single model handles speech recognition, speech translation to English, spoken language identification, and voice activity detection. It ships in six sizes (tiny, base, small, medium, large, turbo — with English-only .en variants for the smaller ones) trading speed against accuracy, and became the foundation of the open transcription ecosystem (whisper.cpp, faster-whisper, and most local transcription tools descend from it).

Install the package (Python 3.8–3.11 supported per README):

```bash

pip install -U openai-whisper

```

Install the ffmpeg command-line tool (required for audio I/O):

```bash

# Ubuntu/Debian

sudo apt update && sudo apt install ffmpeg

# macOS

brew install ffmpeg

# Windows (Chocolatey)

choco install ffmpeg

```

Transcribe from the command line:

```bash

whisper audio.mp3 --model turbo

```

Or from Python (models auto-download on first use):

```python

import whisper

model = whisper.load_model("turbo")

result = model.transcribe("audio.mp3")

print(result["text"])

```

Tip: if pip install fails on tiktoken, you may need a Rust toolchain (setuptools-rust) on your platform — documented in the README.