Live audio
translation API.
Built for your app.

Live audio translation converts speech during a call into translated speech for the listener. CallPayMin also shows translated captions, so you can hear and read the conversation in your selected language. It is automated translation, not a human interpreter.

Yes. Choose captions only if you prefer to hear the original voice and read the translation. Choose live audio + captions to hear translated speech and follow the text. Both participants must agree before translation starts.

The current translated audio output list includes 42 languages, including English, Spanish, German, Polish, Hindi, Kannada, Vietnamese and Japanese. Choose from the languages offered in the call screen or API capabilities response. Recognition and translation quality can vary by language, accent, background noise and mixed-language speech.

The translation flow supports web-to-web, mobile-to-mobile and web-to-mobile calls. Use an updated app and a working microphone. Browser audio playback may require a tap. A stable connection and headphones help keep speech clear.

Organization translation is included in the subscription within monthly limits, with no separate wallet charge. Audio translation billing starts after both audio connections report ready; captions-only billing starts on acceptance. See the API documentation for settlement details.

Yes. The CallPayMin organization API provides consent controls, participant-scoped gateway credentials, translated audio and captions for active two-person calls. Your backend authenticates users and holds the organization API key. Your web or native client handles microphone capture and playback using the SDK transport.

Translation happens during the conversation, with a processing delay. It is not guaranteed to be instant or error-free. Use short, clear turns and confirm important details. For consequential legal, medical or safety decisions, use a qualified interpreter where needed.

This feature is for two-person audio and video calls. The live translation API does not currently cover group meetings or meeting rooms.

End-to-end latency from speech to translated audio is typically 1.5 to 3 seconds end-to-end, depending on language pair, network conditions, and speech length. Caption rendering is faster than translated speech because captions can stream partial results before the final translation finishes.

Create a CallPayMin organization account and generate a test API key (prefix cpm_test_) to try the full flow without billing a real card. Production keys (cpm_live_) consume your organization’s included monthly translation allowance, with no separate wallet charge. Contact the integration team for extended trials or volume pricing.

Twilio provides call infrastructure and now a real-time transcription product, but you still have to wire translation and translated speech synthesis yourself. AWS Transcribe and Deepgram handle speech-to-text at scale, but translation and translated voice output are separate services you must integrate and bill. The CallPayMin live translation API bundles speech-to-text, machine translation, translated voice, captions, two-party consent, and subscription usage limits into a single API. You keep your call infrastructure; we handle the translation path.

Audio is streamed through the translation gateway for processing and is not retained after the session ends unless your organization opts in to a transcript export. Transcripts and captions belong to the organization that initiated the call. CallPayMin does not share audio, transcripts, or caption text with other customers. Review the privacy policy and data processing addendum for the full retention and subprocessor list.