Simple HTTP server for audio transcription using whisper.cpp with Go bindings.

# Build ./build.sh # Run ./whisper-server

# Build image docker build -t whisper-server . # Run container docker run -p 8080:8080 whisper-server

• WHISPER_MODEL - Model to use (default: base.en)
• SERVER_PORT - Server port (default: 8080)
• MAX_FILE_SIZE - Max file size in bytes (default: 25MB)

• GET / - Server info
• GET /health - Health check
• POST /transcribe - Transcribe audio file (WAV format)

# Upload audio file curl -X POST http://localhost:8080/transcribe \ -F "file=@audio.wav" \ -F "language=auto"