Verified:GitHub

whisper

Transcribes multilingual audio into written transcripts using Whisper models.

2|Updated Apr 25, 2026

Search Tags:#translation #transcription #whisper #multilingual #audio-processing #speech-recognition #language-identification

Ready to run

npx skills add https://github.com/AlexiosBluffMara/mercury --skill whisper-alexiosbluffmara

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: whisper
Source: https://github.com/AlexiosBluffMara/mercury/tree/main/optional-skills/mlops/whisper
Command: npx skills add https://github.com/AlexiosBluffMara/mercury --skill whisper-alexiosbluffmara

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires openai-whisper, transformers, torch.

What problem does it solve?

Whisper provides accurate multilingual speech recognition to convert audio into written transcripts, enabling fast transcription, translation, and accessibility across languages.

Core Features & Use Cases

Transcribes audio into text in 99 languages
Detects language and can translate to English when needed
Offers model sizes from tiny to large to balance speed and accuracy
Suitable for podcasts, meetings, interviews, and multimedia archives

Quick Start

Transcribe an audio file using Whisper to obtain a text transcript.

Frequently Asked Questions about whisper

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I transcribe multilingual audio into written transcripts?▼

To transcribe multilingual audio into written transcripts, use pre-trained Whisper models ranging from tiny to large to process podcasts, meetings, and interviews. This requires Python with the openai-whisper, transformers, and torch packages installed.

Can I detect spoken language and translate audio to English automatically?▼

Yes, you can detect spoken language and translate audio to English automatically. Whisper provides language identification alongside its speech recognition capabilities, detecting the spoken language in audio content and optionally translating the transcription output into English when needed.

Do I need Python and torch to run Whisper for speech recognition?▼

Yes, you need Python and torch to run Whisper for speech recognition. The implementation requires a Python environment with the openai-whisper, transformers, and torch packages installed to execute the pre-trained models for audio processing and language identification.

Which Whisper model size is best for balancing transcription speed and accuracy?▼

To balance transcription speed and accuracy, Whisper offers model sizes from tiny to large. Smaller models prioritize faster audio processing, while larger models deliver higher accuracy for multilingual speech recognition, allowing you to choose based on your specific transcription requirements.

What are the limitations of using Whisper for audio processing?▼

Limitations of using Whisper for audio processing include the dependency on heavy packages like torch and transformers, and the trade-off between model size and processing speed. Selecting larger models for high accuracy in multilingual speech recognition demands significantly more computational resources.