ComfyUI Node
Miso TTS - Whisper Transcribe
Transcribe ComfyUI AUDIO with Whisper and output transcript text for reference_text.
Miso TTS - Whisper Transcribe
- audio
- transcript
◄modelwhisper-large-v3-turbo (auto-download)►
◄dtypeauto►
◄languageauto►
◄tasktranscribe►
◄chunk_length_s30►
◄download_if_missingtrue►
CategoryMiso TTS
Inputs (7)
| Name | Type | Default | Description |
|---|---|---|---|
| audio | AUDIO | Audio to transcribe. Connect this output to Miso reference_text. | |
| model | COMBO | whisper-large-v3-turbo (auto-download) | Whisper ASR model. Auto-download options are stored under ComfyUI/models/audio_encoders/. |
| dtype | COMBO | auto | Whisper precision. auto uses bf16/fp16 on GPU and fp32 on CPU. |
| language | COMBO | auto | Optional language hint for Whisper. |
| task | COMBO | transcribe | transcribe keeps the source language; translate outputs English. |
| chunk_length_s | INT | 300–120 | Chunk length in seconds for longer audio. 0 disables chunking. |
| download_if_missing | BOOLEAN | true | Download selected auto-download Whisper model if missing. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| transcript | STRING | — |