Qwen 3 TTS Text to Speech 1.7B text to speech fal AI

Needs text.

Inputs

1 required

Text: The text to be converted to speechRequired · text
Prompt: Optional prompt to guide the style of the generated speechOptional · prompt
Subtalker dosample: Sampling switch for the sub-talkerOptional · subtalker_dosample
Subtalker temperature: Temperature for sub-talker samplingOptional · subtalker_temperature
Temperature: Sampling temperature; higher => more randomOptional · temperature
Max new tokens: Maximum number of new codec tokens to generateOptional · max_new_tokens
Voice: The voice to be used for speech synthesis, will be ignored if a speaker embedding is providedOptional · voice
Subtalker top p: Top-p for sub-talker samplingOptional · subtalker_top_p
Subtalker top k: Top-k for sub-talker samplingOptional · subtalker_top_k
Language: The language of the voiceOptional · language
Repetition penalty: Penalty to reduce repeated tokens/codesOptional · repetition_penalty
Enable safety checker: Authorized callers may set false to request relaxed screeningOptional · enable_safety_checker
Top k: Top-k sampling parameterOptional · top_k
Top p: Top-p sampling parameterOptional · top_p

Your agent reads the schema and the price with inspect_tool before it runs the tool.

AI audio