skills.rest
erfianugrah
erfianugrahCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

whisper

Transcribe audio and video via a local WhisperX HTTP API.

1|Updated Mar 18, 2024
Search Tags:#translation#youtube#transcription#whisperx#speech-to-text#diarization#http api
One-click install
npx skills add https://github.com/erfianugrah/dotfiles --skill whisper-erfianugrah
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: whisper
Source: https://github.com/erfianugrah/dotfiles/tree/main/.pi/agent/skills/whisper
Command: npx skills add https://github.com/erfianugrah/dotfiles --skill whisper-erfianugrah

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Transcribing long audio/video into readable text is slow and error-prone when done manually, especially for podcasts, lectures, and YouTube content.

Core Features & Use Cases

  • YouTube to transcript: Download a YouTube media file on the whisper server and then transcribe it into plain text, suitable for creating summaries or searchable notes.
  • Local file transcription: Transcribe audio/video files available on the whisper server filesystem using WhisperX models.
  • Async job handling + GPU orchestration: Submit transcription as an async job, poll for completion, and rely on the llm-compose stack GPU swap behavior to avoid GPU contention.

Quick Start

Use the whisper skill to transcribe a YouTube link by first downloading it via the whisper HTTP API, then submitting the returned server-side file path to the transcribe endpoint and waiting for the job to finish.

Frequently Asked Questions about whisper

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

FAQPage Schema
How do I transcribe a YouTube video to text using speech-to-text?▼

To transcribe a YouTube video to text, the Skill downloads the media via a local WhisperX HTTP API endpoint and submits the server-side file path to an asynchronous transcription API, returning plain text upon job completion.

Can I use diarization and hotwords for audio transcription?▼

Yes, the transcription API supports diarization, hotwords, language hints, translation, and initial prompt context to improve accuracy when converting audio to text.

Do I need a local server to run WhisperX transcription jobs?▼

Yes, you need a locally hosted WhisperX HTTP API running at http://localhost:7860 or a specified WHISPER_URL, which handles the audio download and asynchronous transcription workload.

How does asynchronous job handling work for long audio transcription?▼

Long audio transcription submits jobs to the /api/transcribe endpoint and polls the /api/jobs/<job_id> status endpoint until processing finishes, preventing timeouts during lengthy speech-to-text operations.

What is the best way to avoid GPU contention during video transcription?▼

To avoid GPU contention during video transcription, the system relies on the llm-compose stack GPU swap behavior to manage resources while polling asynchronous transcription jobs for completion.

Related Skills

Explore More Content & Communication Skills→
steipetesteipete
385.7k

sag

Convert text to natural-sounding speech via the sag CLI with ElevenLabs TTS.

Community
Intermediate
mindmorassmindmorass
2

transcript-summarizer

Converts meeting transcripts into structured summaries highlighting decisions, actions, and topics.

Community
Advanced
shige1014-devshige1014-dev

bfl-api

Integrate BFL FLUX API endpoints for image generation, editing, and async retrieval.

Community
Advanced
framerslabframerslab

google-cloud-stt

Transcribe audio files in batches via the Google Cloud Speech-to-Text API.

Official
Intermediate