Audio Transcribe

A Go-based audio transcription tool that uses Google Cloud Platform's Vertex AI Gemini to transcribe audio files and generate summaries with timecoded speaker identification.

Quickstart

Prerequisites

1. Google Cloud Setup

   # Install gcloud CLI and authenticate
   gcloud auth application-default login
   
   # Set your project ID
   export GCP_PROJECT="your-project-id"

2. Install Dependencies

   • Go 1.19+
   • ffmpeg (for audio splitting)

Installation

1. Clone and build:

   git clone https://github.com/owulveryck/audiotranscribe.git
   cd audiotranscribe
   go build -o audiotranscribe .

Usage

Single audio file:

./audiotranscribe audio.m4a

Multiple audio files with output file:

./audiotranscribe -o transcript.md audio1.m4a audio2.m4a

Large files (auto-split into 25min chunks):

./split_and_transcribe.sh large_audio.m4a

Environment Variables

• GCP_PROJECT (required) - Your Google Cloud project ID
• GEMINI_MODEL (optional) - Gemini model to use (default: "gemini-2.0-flash")
• GCP_REGION (optional) - GCP region (default: "europe-west9")

Output

The tool generates markdown files with:

• Timestamped transcripts with speaker identification
• Combined summaries for multiple files
• Structured format for easy reading

Example output placed in same directory as input files.