Icana Transcription Service

Copy as Markdown

The Icana Transcription Service is an API for accurate Australian-English speech-to-text transcription. Built for Australian accents, terminology, and business contexts, powered by Whisper and pyannote.

Key Features

• Australian-optimised — Tuned for Australian English, including local accents, place names, and terminology
• Speaker diarisation — Identify and label different speakers in a conversation
• Flexible language support — Default is English, with support for additional Whisper languages
• Prompt guidance — Provide context prompts to improve transcription accuracy
• Simple REST API — Upload, transcribe, poll for results, and clean up

How It Works

1. Upload your audio file via the /upload endpoint
2. Submit the transcription job via /transcribe with the returned S3 URI
3. Poll the /status/{job_id} endpoint until the job completes
4. Retrieve the transcription text and diarisation output from the status response
5. Clean up with the /delete/{job_id} endpoint when done

Getting Started

1. Get your API key
2. Follow the Quickstart guide to transcribe your first file
3. Explore the full API Reference

SDKs

Official client libraries are available for:

• Python
• JavaScript / Node.js

Support

• Email: support@icana.ai
• Website: icana.ai