Openai Whisper

# Openai Whisper Transcribe audio files locally using OpenAI's Whisper speech-to-text model -- no API key required, with support for multiple languages and output formats. The OpenAI Whisper skill enables your AI assistant to transcribe audio files to text entirely on your local machine. It uses OpenAI's open-source Whisper model to convert speech from audio files (MP3, M4A, WAV, etc.) into text, subtitles, or other output formats. It also supports translation of non-English audio to English text. No API key or internet connection is needed after the initial model download. ## How It Works The skill uses the whisper CLI, which runs the Whisper neural network locally on your machine. On first use, the selected model is downloaded to ~/.cache/whisper. You provide an audio file path, choose a model size (balancing speed vs accuracy), and specify an output format. The CLI processes the audio and outputs the transcription to the specified directory. ## Key Features Local Processing — All transcription runs on your machine with no API key or cloud dependencyMultiple Model Sizes — Choose from tiny to large models depending on your speed vs. accuracy needs (defaults to turbo)Translation Support — Translate non-English audio directly to English text using the --task translate flagMultiple Output Formats — Output as plain text (txt), SubRip subtitles (srt), and other formatsBroad Audio Format Support — Works with MP3, M4A, WAV, and other common audio formats ## Requirements Whisper CLI — The OpenAI Whisper package must be installed. Installable via Homebrew ## Use Cases Meeting Transcription — Convert recorded meetings or calls into searchable text documentsSubtitle Generation — Generate SRT subtitle files from video or audio recordingsContent Creation — Transcribe interviews, podcasts, or lectures for written contentMultilingual Translation — Translate foreign-language audio recordings into English text ## Installation Install via: `npx clawhub@latest install openai-whisper`

Instalação

Habilidades relacionadas