transcribe

name	transcribe
description	Transcribe audio files with a local or self-hosted speech-to-text endpoint, with Gemini fallback when configured. Use for voice messages, audio attachments, or explicit transcription requests.

Audio Transcription

Transcribe audio files with a speech-to-text server that exposes an OpenAI-style transcription API.

Configuration

Set TRANSCRIBE_API_URL to your speech-to-text endpoint.

• default example: http://127.0.0.1:10301
• expected endpoint: <base-url>/v1/audio/transcriptions

Optional fallback:

• GEMINI_API_KEY
• GOOGLE_API_KEY

Quick Start

./scripts/transcribe.sh /path/to/audio.ogg
./scripts/transcribe.sh /path/to/audio.ogg en

Manual Curl

curl -s -X POST "${TRANSCRIBE_API_URL:-http://127.0.0.1:10301}/v1/audio/transcriptions" \
  -F "file=@/path/to/audio.ogg" \
  -F "model=whisper-1"

Supported Formats

• .ogg
• .wav
• .mp3
• .m4a
• .webm

Health Check

curl -s "${TRANSCRIBE_API_URL:-http://127.0.0.1:10301}/health"

Notes

• the endpoint values in this repository are placeholders
• replace them with your own local or hosted service
• if Gemini fallback is enabled, the script can still return a transcription when the local endpoint is unavailable