Skip to content
Get packs

Whisper Speech Transcription

Transcribe and translate speech in 99 languages using OpenAI's Whisper model, with subtitles and timestamps.

Source Download skill
Repository
Orchestra-Research/AI-Research-SKILLs
Stars
14,195
Licence
MIT
Files
2 files

What it does

Once installed, the agent knows how to use OpenAI's Whisper speech recognition model to turn audio or video recordings into written text. It can pick a model size based on speed and accuracy needs, translate non-English speech into English, add word-level timestamps, and export subtitles in SRT, VTT, TXT or JSON formats. It also includes guidance for batch processing many files and for using faster-whisper for quicker results.

Great for

  • Creators who need subtitles or transcripts for podcasts and videos
  • Developers building a multilingual speech-to-text step into an app or pipeline
  • Students or professionals turning meeting or lecture recordings into searchable notes

Skip it if

  • Highly sensitive recordings if you cannot send or store audio with a third party; the skill itself runs locally, but any cloud use you add would send data out (the skill lists AssemblyAI, Deepgram and Google as alternatives that are managed services)
  • Very long recordings without splitting, since the skill notes accuracy degrades past about 30 minutes and the model can hallucinate text
  • Real-time live captioning, since the skill's main approach processes files and streaming is only briefly covered via faster-whisper

Try asking

  • “Transcribe this podcast episode file into text with timestamps”
  • “Translate this Spanish interview recording into English text”
  • “Generate an SRT subtitle file for my video called demo.mp4”
  • “Batch transcribe all the .mp3 files in this folder into .txt files”
Step-by-step guide: how to use Whisper Speech TranscriptionInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the MIT licence and an attribution note.

    whisper.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip whisper.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip whisper.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use the turbo model for a good balance of speed and quality, and specify the language when you know it to speed things up.
  • Add an initial prompt with names and technical terms so the transcript spells them correctly.
  • Split long recordings into chunks of about 30 minutes or less and check outputs for invented text.

What's inside

  • references/languages.md
  • SKILL.md

Redistributed unmodified from Orchestra-Research/AI-Research-SKILLs under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.