speech-to-text · noizai — Claude Code skill: convert audio to text
speech-to-text converts audio and video files to text with automatic language detection, timestamps, and speaker identification. Requires a Noiz API key.
no license 198 installs ★ 539 updated 2026-09-28
npx skills add NoizAI/skills --skill speech-to-text -a claude-codeWhat speech-to-text does
The skill runs a Python script that uploads audio files to Noiz servers for processing. It handles files up to 50 MB and 10 minutes long in mp3, wav, m4a, ogg, flac, aac, and webm formats. By default it outputs plain text, but can return full JSON with segment breakdowns, timing information, and speaker labels.
You need the requests package (install via pip) and a Noiz API key to proceed. Save the key once through configuration or pass it as the NOIZ_API_KEY environment variable. Pricing is $0.0006 per audio second; a ten-minute file runs about $0.36.
The skill activates on commands like "transcribe", "get subtitles", "identify speakers" or "what does this audio say". It supports automatic multilingual detection and explicit language selection via the --language flag.
What it helps with
- transcribe a podcast or interview into searchable text
- generate subtitles with timings from a video recording
- separate and label speech by different speakers in audio
- recognize and convert speech in multiple languages
- export transcription as JSON with segment metadata
Example prompts for the agent
- Transcribe
meeting.wavand save the result totranscript.txt - Get a transcript of
podcast.m4ain Spanish with speaker labels output as JSON - Convert
interview.mp3to text with automatic language detection
How to install speech-to-text
1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill speech-to-text -a claude-code2Manually: put the files in place
~/.claude/skills/speech-to-text/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/speech-to-text/
1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill speech-to-text -a codex2Manually: put the files in place
.agents/skills/speech-to-text/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill speech-to-text -a cursor2Manually: put the files in place
.cursor/skills/speech-to-text/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install NoizAI/skills skills/speech-to-text2Manually: put the files in place
.github/skills/speech-to-text/SKILL.md1Skills CLI (Vercel)
npx skills add NoizAI/skills --skill speech-to-text -a gemini-cli2Platform CLI
gemini skills install https://github.com/NoizAI/skills --path skills/speech-to-textManually: put the files in place
.devin/skills/speech-to-text/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add NoizAI/skills --skill speech-to-text -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about speech-to-text
How do I install speech-to-text in Claude Code?
Run npx skills add NoizAI/skills --skill speech-to-text -a claude-code in your terminal. After installation, the skill will be available at ~/.claude/skills/speech-to-text/SKILL.md.
What do I need to use speech-to-text?
Install the Python requests package with pip install requests and get a Noiz API key. Retrieve your key from developers.noiz.ai, save it via python3 skills/speech-to-text/scripts/stt.py config --set-api-key YOUR_KEY, or set the NOIZ_API_KEY environment variable.
How can I get timestamps and speaker identification in the output?
Use the --json flag: python3 skills/speech-to-text/scripts/stt.py audio.wav --json -o result.json. The output will include segments with text, start and end times, and speaker numbers.
Similar skills
All skills for AI and ML →speech-to-text by other authors comes first.