feat: add Voice Design node + language and guidance_scale to Generate

OmniVoiceVoiceDesign: structured dropdowns for gender/age/pitch/accent
that compose into an instruct string — wire to Generate's instruct input.

OmniVoiceGenerate: new optional language dropdown (auto + 11 languages)
and guidance_scale (CFG, default 2.0) parameters.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

__init__.py

@@ -1,4 +1,4 @@
-from .nodes import OmniVoiceModelLoader, OmniVoiceGenerate, OmniVoiceEpubLoader, OmniVoiceVoicePreset, OmniVoiceMixVoices
+from .nodes import OmniVoiceModelLoader, OmniVoiceGenerate, OmniVoiceEpubLoader, OmniVoiceVoicePreset, OmniVoiceMixVoices, OmniVoiceVoiceDesign
 
 NODE_CLASS_MAPPINGS = {
     "OmniVoiceModelLoader": OmniVoiceModelLoader,
@@ -6,6 +6,7 @@ NODE_CLASS_MAPPINGS = {
     "OmniVoiceEpubLoader": OmniVoiceEpubLoader,
     "OmniVoiceVoicePreset": OmniVoiceVoicePreset,
     "OmniVoiceMixVoices": OmniVoiceMixVoices,
+    "OmniVoiceVoiceDesign": OmniVoiceVoiceDesign,
 }
 
 NODE_DISPLAY_NAME_MAPPINGS = {
@@ -14,6 +15,7 @@ NODE_DISPLAY_NAME_MAPPINGS = {
     "OmniVoiceEpubLoader": "OmniVoice EPUB Loader",
     "OmniVoiceVoicePreset": "OmniVoice Voice Preset",
     "OmniVoiceMixVoices": "OmniVoice Mix Voices",
+    "OmniVoiceVoiceDesign": "OmniVoice Voice Design",
 }
 
 __all__ = ["NODE_CLASS_MAPPINGS", "NODE_DISPLAY_NAME_MAPPINGS"]

nodes/__init__.py

@@ -3,5 +3,6 @@ from .generator import OmniVoiceGenerate
 from .epub_loader import OmniVoiceEpubLoader
 from .voice_presets import OmniVoiceVoicePreset
 from .mix_voices import OmniVoiceMixVoices
+from .voice_design import OmniVoiceVoiceDesign
 
-__all__ = ["OmniVoiceModelLoader", "OmniVoiceGenerate", "OmniVoiceEpubLoader", "OmniVoiceVoicePreset", "OmniVoiceMixVoices"]
+__all__ = ["OmniVoiceModelLoader", "OmniVoiceGenerate", "OmniVoiceEpubLoader", "OmniVoiceVoicePreset", "OmniVoiceMixVoices", "OmniVoiceVoiceDesign"]

nodes/generator.py

@@ -62,6 +62,14 @@ class OmniVoiceGenerate:
                     "default": "",
                     "tooltip": "Transcription of ref_audio. Connect a Whisper (or other STT) node for best results.",
                 }),
+                "language": (
+                    ["auto", "English", "Chinese", "Japanese", "Korean", "French",
+                     "Spanish", "German", "Portuguese", "Russian", "Arabic", "Hindi"],
+                    {
+                        "default": "auto",
+                        "tooltip": "Language of the text to synthesize. 'auto' lets the model detect it.",
+                    },
+                ),
                 "instruct": ("STRING", {
                     "default": "",
                     "tooltip": (
@@ -79,6 +87,13 @@ class OmniVoiceGenerate:
                         "EXAMPLE:  female, high pitch, british accent"
                     ),
                 }),
+                "guidance_scale": ("FLOAT", {
+                    "default": 2.0, "min": 0.0, "max": 20.0, "step": 0.1,
+                    "tooltip": (
+                        "Classifier-free guidance scale. Higher = more faithful to the reference/instruct, "
+                        "but can over-saturate. 2.0 is a good default."
+                    ),
+                }),
                 "speed": ("FLOAT", {
                     "default": 1.0, "min": 0.1, "max": 3.0, "step": 0.1,
                     "tooltip": "Playback speed multiplier. 1.0 = normal, >1.0 = faster, <1.0 = slower.",
@@ -104,10 +119,13 @@ class OmniVoiceGenerate:
     FUNCTION = "generate"
     CATEGORY = "OmniVoice"
 
-    def generate(self, model, text, mode, ref_audio=None, ref_text="", instruct="", speed=1.0, num_step=32, seed=0):
+    def generate(self, model, text, mode, ref_audio=None, ref_text="", language="auto",
+                 instruct="", guidance_scale=2.0, speed=1.0, num_step=32, seed=0):
         if seed != 0:
             torch.manual_seed(seed)
-        kwargs = {"text": text, "speed": speed, "num_step": num_step}
+        kwargs = {"text": text, "speed": speed, "num_step": num_step, "guidance_scale": guidance_scale}
+        if language != "auto":
+            kwargs["language"] = language
 
         if mode == "voice_cloning" and ref_audio is None:
             raise ValueError("voice_cloning mode requires ref_audio to be connected")

nodes/voice_design.py

@@ -0,0 +1,32 @@
+class OmniVoiceVoiceDesign:
+    """Compose a voice design instruct string from structured dropdowns."""
+
+    GENDERS  = ["none", "male", "female"]
+    AGES     = ["none", "child", "teenager", "young adult", "middle-aged", "elderly"]
+    PITCHES  = ["none", "very low pitch", "low pitch", "moderate pitch", "high pitch", "very high pitch", "whisper"]
+    ACCENTS  = [
+        "none",
+        "american accent", "british accent", "australian accent", "canadian accent",
+        "indian accent", "chinese accent", "japanese accent", "korean accent",
+        "portuguese accent", "russian accent",
+    ]
+
+    @classmethod
+    def INPUT_TYPES(cls):
+        return {
+            "required": {
+                "gender": (cls.GENDERS, {"default": "female"}),
+                "age":    (cls.AGES,    {"default": "none"}),
+                "pitch":  (cls.PITCHES, {"default": "none"}),
+                "accent": (cls.ACCENTS, {"default": "none"}),
+            },
+        }
+
+    RETURN_TYPES  = ("STRING",)
+    RETURN_NAMES  = ("instruct",)
+    FUNCTION      = "compose"
+    CATEGORY      = "OmniVoice"
+
+    def compose(self, gender, age, pitch, accent):
+        parts = [v for v in [gender, age, pitch, accent] if v != "none"]
+        return (", ".join(parts),)