Paul de LavallazandClaude Opus 4.6 5f589d0dad Add AUDIO output for ComfyUI native audio pipeline compatibility
Enables direct connection to nodes that accept AUDIO input (e.g. Fabric 1.0 Lipsync).
Converts the generated WAV/MP3 to a waveform tensor + sample_rate dict.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-12 00:47:03 -03:00
2025-11-04 23:56:11 -03:00
2025-11-04 23:56:11 -03:00
2025-11-04 23:56:11 -03:00
2025-11-04 23:56:11 -03:00
2025-11-04 23:56:11 -03:00

ComfyUI-CartesiaTTS

Nodo para Cartesia Sonic-3 que llama al endpoint /tts/bytes y devuelve:

  • file_path: ruta absoluta del archivo (wav/mp3/raw)
  • bytes: audio en binario
  • url: file://... o un link subido a tmpfiles.org (si activás upload_to_tmpfiles)

Instalación

  1. Cloná este repo dentro de ComfyUI/custom_nodes/ComfyUI-CartesiaTTS/
  2. Reiniciá ComfyUI.

Uso (inputs clave)

  • api_key: tu Cartesia API Key
  • transcript: texto a sintetizar
  • voice_id: ID de voz
  • Opcionales: container (wav/mp3/raw), encoding (pcm_f32le, etc.), sample_rate (44100), gen_speed, gen_volume
  • upload_to_tmpfiles: si se activa, intenta subir el archivo y devolver un link público (best-effort).

Flujo típico

[External Text transcript] → [Cartesia Sonic-3 TTS]
                       file_path / url / bytes
                 → (url) HeyGen /video/generate (voice.type="audio", audio_url=url)
                 → Poll → Download video

Requisitos

  • requests

Licencia

MIT

S
Description
No description provided
Readme MIT
31 KiB
Languages
Python 100%