ComfyUI Node
MOSS-TTS Voice Clone
Zero-shot voice cloning from a ComfyUI AUDIO reference.
MOSS-TTS Voice Clone
- mosstts_model
- reference_audio
- audio
◄textThis line will be spoken in the reference voice.►
◄languageauto►
◄instruction►
◄duration_tokens0►
◄quality►
◄sound_event►
◄ambient_sound►
◄max_new_tokens4096►
◄do_sampletrue►
◄text_temperature1.00►
◄text_top_p1.00►
◄text_top_k50►
◄audio_temperature1.70►
◄audio_top_p0.80►
◄audio_top_k25►
◄audio_repetition_penalty1.00►
◄seed0►
CategoryMOSS-TTS
Inputs (19)
| Name | Type | Default | Description |
|---|---|---|---|
| mosstts_model | MOSSTTS_MODEL | Loaded MOSS-TTS model bundle. | |
| reference_audio | AUDIO | Reference speaker audio. The node encodes it with MOSS-Audio-Tokenizer-v2 before generation. | |
| text | STRING | This line will be spoken in the reference voice. | Text to synthesize. Inline [pause 0.5s], Pinyin, and IPA are passed through to MOSS-TTS. |
| language | COMBO | auto | Language hint. v1.5 performs best when the language is specified. |
| instruction | STRING | Optional free-form style or delivery instruction. | |
| duration_tokens | INT | 00–45000 | Optional duration target in audio tokens. 0 omits the field. MOSS runs around 12.5 tokens per second. |
| quality | STRING | Optional quality hint field exposed by the MOSS user-message schema. | |
| sound_event | STRING | Optional sound-event hint field exposed by the MOSS user-message schema. | |
| ambient_sound | STRING | Optional ambient-sound hint field exposed by the MOSS user-message schema. | |
| max_new_tokens | INT | 40961–45000 | Generation budget in audio frames. At 12.5 frames per second, 4096 is roughly 5.5 minutes. |
| do_sample | BOOLEAN | true | Use stochastic sampling. Disable for deterministic greedy decoding. |
| text_temperature | FLOAT | 1.000–2 | Sampling temperature for assistant text/audio-control tokens. |
| text_top_p | FLOAT | 1.000–1 | Nucleus sampling for assistant text/audio-control tokens. |
| text_top_k | INT | 500–4096 | Top-K sampling for assistant text/audio-control tokens. |
| audio_temperature | FLOAT | 1.700–3 | Recommended v1.5 audio sampling temperature. |
| audio_top_p | FLOAT | 0.800–1 | Recommended v1.5 nucleus sampling cutoff for audio codebooks. |
| audio_top_k | INT | 250–1024 | Recommended v1.5 Top-K cutoff for audio codebooks. |
| audio_repetition_penalty | FLOAT | 1.001–2 | Penalty for repeated acoustic code patterns. |
| seed | INT | 00–9223372036854776000 | 0 leaves sampling unseeded. Positive values make identical settings repeatable. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| audio | AUDIO | — |