ComfyUI Node
Audio Transcript Batch (CRT)
Transcribes a batched AUDIO input with faster-whisper (CTranslate2). Outputs one transcript per batch item as a list that pairs item by item with the batch loader's file_names in SaveTextWithPath.
Audio Transcript Batch (CRT)
- audio
- strings_batch
- display
- status
◄isolate_voicefalse►
◄keep_model_loadedtrue►
◄whisper_modellarge-v3-turbo►
◄compute_typefloat16►
◄vad_filtertrue►
◄transcribe_batch_size8►
◄beam_size5►
◄languageauto►
CategoryCRT/Audio
Inputs (9)
| Name | Type | Default | Description |
|---|---|---|---|
| audio | AUDIO | — | |
| isolate_voice | BOOLEAN | false | Run MelBandRoFormer voice isolation on each file before transcription. |
| keep_model_loaded | BOOLEAN | true | Keep the Whisper model loaded in VRAM after batch transcription. Disable to free VRAM immediately. |
| whisper_modelopt | COMBO | large-v3-turbo | 14 options: tiny.en, tiny, base.en, base, small.en, small, +8 |
| compute_typeopt | COMBO | float16 | 4 options: float16, int8_float16, int8, float32 |
| vad_filteropt | BOOLEAN | true | Cuts silences before inference for faster long-audio processing. Uses Silero VAD to skip silent regions. |
| transcribe_batch_sizeopt | INT | 81–64 | Number of 30s chunks decoded in parallel per file. 1 disables batched decoding. |
| beam_sizeopt | INT | 51–10 | Beam search width. 1 = greedy decoding (faster, slightly less accurate). |
| languageopt | STRING | auto | — |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| strings_batch | STRING | One transcript per batch item as a list — pairs item by item with the batch loader's file_names in SaveTextWithPath. |
| display | STRING | Same transcripts stitched into one display string, separated by a blank line per item. |
| status | STRING | Summary status for the batch run. |