Camb.ai icon

Camb.ai

Use Camb.ai for TTS, transcription, translation, voice cloning, and more

Actions9

Audio → Synthesize (TTS)

AI-generated

Summary

Convert text to speech using Camb.ai's MARS TTS models, returning a binary audio file and metadata.

Inputs

  • text (required) — Text to convert to speech (3-3000 characters)
  • voiceId (required) — Voice ID to use for synthesis (integer)
  • language (required) — BCP-47 language code for synthesis
  • model — The MARS TTS model to use (mars-flash: 22.05kHz; mars-pro: 48kHz; mars-instruct: 22.05kHz, supports user instructions)
  • outputFormat — Audio output format (wav, flac, adts, pcm_s16le)
  • binaryPropertyName (required) — The name of the output binary field to put the audio file in
  • options — Optional collection: userInstructions (3-1000 chars, requires mars-instruct), enhanceNamedEntities (boolean), fileName (custom filename)

Output shape

object with binary and JSON fields

Binary audio data written to the field specified by binaryPropertyName; JSON includes text, voiceId, language, model, outputFormat, fileName, and size (byte length of audio buffer). For pcm_s16le output, a WAV header is prepended to the raw PCM data.

Links

Discussion