ComfyUI Node
Spark-TTS 音色克隆（AICG3D 内置兼容）
首次运行前，请把完整 Spark-TTS-0.5B 模型放到 ComfyUI/models/tts/Spark-TTS-0.5B。下载地址：https://huggingface.co/SparkAudio/Spark-TTS-0.5B
Spark-TTS 音色克隆（AICG3D 内置兼容）
- custom_clone_audio
- audio
◄text►
◄cloned_speaker婷婷RAP►
◄temperature0.80►
◄top_k50►
◄top_p0.95►
◄max_new_tokens3000►
◄do_sampletrue►
◄unload_modeltrue►
◄seed0►
◄custom_clone_text►
CategoryAICG3D/提示词与 Skill
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| text | STRING | 首次运行前，请把完整 Spark-TTS-0.5B 模型放到 ComfyUI/models/tts/Spark-TTS-0.5B。下载地址：https://huggingface.co/SparkAudio/Spark-TTS-0.5B | |
| cloned_speaker | COMBO | 婷婷RAP | 5 options: 婷婷RAP, 婷婷哼唱, 婷婷, 明文, 撒娇小美 |
| temperature | FLOAT | 0.800–1.5 | — |
| top_k | INT | 50 | — |
| top_p | FLOAT | 0.950–1 | — |
| max_new_tokens | INT | 3000 | — |
| do_sample | BOOLEAN | true | — |
| unload_model | BOOLEAN | true | — |
| seed | INT | 00–18446744073709550000 | — |
| custom_clone_textopt | STRING | — | |
| custom_clone_audioopt | AUDIO | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| audio | AUDIO | — |