stt

Transcribe audio recordings to text via a Whisper-compatible STT backend.

2|2|Updated Feb 22, 2026

Search Tags:#api #transcription #ffmpeg #speech-to-text #stt #audio-processing

Ready to run

npx skills add https://github.com/mxzinke/atlas --skill stt-mxzinke

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: stt
Source: https://github.com/mxzinke/atlas/tree/main/app/defaults/skills/stt
Command: npx skills add https://github.com/mxzinke/atlas --skill stt-mxzinke

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ffmpeg, ffprobe, curl, python3, bc, pyyaml, and includes scripts (resource) components.

What problem does it solve?

Transcribe audio to text from recordings to enable searchable transcripts and improved accessibility.

Core Features & Use Cases

Supports common audio formats (wav, mp3, ogg, m4a, aac, flac) and converts them into text via a Whisper-compatible STT backend.
Handles long recordings by chunking into 120-second segments with overlaps to preserve context.
Language hint support and automatic URL-based source input for flexible transcription workflows.

Quick Start

Transcribe a local audio file to text with the stt tool.

Frequently Asked Questions about stt

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

FAQPage Schema

How do I transcribe audio to text from common formats like mp3 or m4a?▼

To transcribe audio to text, this Skill converts common formats like mp3, m4a, and wav to 16kHz mono WAV before processing them through a Whisper-compatible backend.

How does speech-to-text handle long recordings like podcasts or meetings?▼

Speech-to-text for long recordings is handled by chunking audio into 120-second segments with overlaps to preserve context during transcription.

Do I need ffmpeg installed to transcribe audio files?▼

Yes, you need ffmpeg installed to transcribe audio files, as it is required to convert various input formats into 16kHz mono WAV prior to transcription.

Can I provide a language hint when transcribing voice notes?▼

Yes, you can provide language hints when transcribing voice notes to improve transcription accuracy through the Whisper-compatible STT backend.

What is the best way to transcribe audio from a URL source?▼

The best way to transcribe audio from a URL source is using the automatic URL-based input feature, which fetches the remote audio for conversion and transcription.

Why does my audio transcription require curl and python3 dependencies?▼

Audio transcription requires curl and python3 dependencies to fetch URL-based sources and execute the scripts that manage the chunking and Whisper-compatible API processing.