| name | aima-asr |
| description | Transcribe audio files using AIMA's current local ASR model (speech-to-text). |
| metadata | {"openclaw":{"emoji":"🎙️","requires":{"bins":["curl"]},"always":true}} |
AIMA Speech-to-Text
Transcribe audio files using the ASR model currently managed by AIMA/OpenClaw.
Quick start
{baseDir}/scripts/transcribe.sh /path/to/audio.wav
Useful flags
{baseDir}/scripts/transcribe.sh /path/to/audio.wav --out /tmp/transcript.txt
{baseDir}/scripts/transcribe.sh /path/to/audio.m4a --json
Output
- Prints transcript to stdout by default
- With
--out, saves to file and prints the path
Notes
- Model: auto-detected from
~/.openclaw/openclaw.json (override with AIMA_ASR_MODEL)
- Supported formats: wav, mp3, m4a, ogg, flac
- Runs on AIMA proxy at
http://127.0.0.1:6188/v1