Krisp Voice Translation API
Real-time speech-to-speech voice translation delivered over a persistent WebSocket session. A REST call mints a short-lived session key, the client opens a WSS session with a JSON config message (source/target BCP-47 language, voice, custom vocabulary, translation dictionary), then streams 16 kHz mono PCM or Opus frames; the server streams synthesized translated speech back on the same socket alongside JSON transcript, translation, and error events. Krisp advertises 61 languages and a 99.9% uptime SLA.
Documentation
Documentation
https://sdk-docs.krisp.ai/docs/voice-translation.md
APIReference
https://sdk-docs.krisp.ai/docs/voice-translation-api.md
GettingStarted
https://sdk-docs.krisp.ai/docs/quick-start.md
Specifications
OpenAPI
https://raw.githubusercontent.com/api-evangelist/krisp/refs/heads/main/openapi/krisp-developers-openapi.yml
AsyncAPI
https://raw.githubusercontent.com/api-evangelist/krisp/refs/heads/main/asyncapi/krisp-voice-translation-asyncapi.yml