qwen3-tts

name	Qwen3-TTS
description	Text-to-speech generation using Qwen3-TTS via FAL API
triggers	["Generate speech audio from dialogue","Multi-character voice synthesis","TTS for manga dialogue"]

Qwen3-TTS Skill

Generate high-quality speech audio from text using Qwen3-TTS.

Usage

from skills.qwen_tts import QwenTTS

tts = QwenTTS()

# Single line
result = await tts.generate_speech("Hello world!")
# result.audio_path, result.duration_seconds

# Multi-character dialogue
results = await tts.generate_dialogue([
    ("Mochi", "Hi there! What's that?"),
    ("Hero", "It's a treasure map!"),
])

API

QwenTTS.generate_speech()

async def generate_speech(
    text: str,
    output_path: Path = None,
    voice_embedding: Optional[Path] = None,
    style_prompt: Optional[str] = None,
) -> TTSResult

QwenTTS.generate_dialogue()

async def generate_dialogue(
    dialogue_lines: list[tuple[str, str]],
    character_voices: dict[str, Path] = None,
    output_dir: Path = None,
) -> list[TTSResult]

Voice Embeddings

Voice embeddings can be created using the FAL voice cloning API. Pass a .safetensors file path to voice_embedding parameter.

Dependencies

• FAL_KEY in .env
• requests
• ffprobe (for duration detection)