Docs/Get started/Quickstart
Quickstart
Authenticate, send one minute of audio, and get a transcription with translation in under 200 milliseconds. This guide takes about 3 minutes.
API · WebSocket · REST
1. Get an API key
Create an API key from your Varmir dashboard. Keep it secret — anyone with the key can spend your credits. Keys are prefixed with mstk_ so they are easy to spot in logs.
2. Make your first call
Send an audio file to the /transcribe endpoint with your key in the Authorization header. Set translate_to to get a translation alongside the transcript.
curl https://api.varmir.com/api/v1/transcribe \
-H "Authorization: Bearer mstk_your_key" \
-F "[email protected]" \
-F "translate_to=uk"0.4 credits used per second of audio.
3. Stream in real-time
For live captions, open a WebSocket and push raw audio chunks. Text comes back as you speak — recognized first, then translated phrase by phrase.
stream.js
const ws = new WebSocket("wss://api.varmir.com/api/v1/stream");
// Authenticate, then stream raw PCM chunks
ws.onopen = () => ws.send(JSON.stringify({ token: "mstk_your_key" }));
ws.onmessage = (e) => console.log(JSON.parse(e.data).text);