The core purpose of VoiceSmiths is to turn a Chinese or English story into a multilingual multi-character audio drama with one click, aimed at content creation scenarios such as short drama going overseas, novel promotion posts, AI comic dramas, and audiobooks. Users do not need to record, find voice actors, or manually edit; the original text can be used directly to generate audio works without modification.
Main capabilities include: automatic role assignment, able to identify the narrator and each character in the story and assign voices; line-by-line emotion annotation, giving each line of dialogue the corresponding emotional expression; providing 54 AI voice actors, covering types such as domineering CEO, queen, Monkey King, and storyteller, with suitable voices automatically selected during role assignment. In terms of multilingual support, it supports English, Spanish, Portuguese, Japanese, Korean, Indonesian, and others, and can adapt into multiple language versions with one click, performed by the same cast of actor voices. The generated results can output MP3, SRT subtitles, and vertical-screen video, with each file containing a built-in AI-generated label and a line-by-line subtitle timeline. The underlying technology is based on ElevenLabs Eleven v4.
The usage process is divided into three steps: Step one, paste the story. Novel chapters, promotion post copy, or scripts are all acceptable, in either Chinese or English, with a maximum of 8,000 characters at a time; Step two, confirm the actors, assigning one voice each for the narrator and each character and marking the emotion line by line. This step is free; Step three, audition and generate. First listen to the beginning, then generate the whole chapter by word count. Credits are deducted according to the number of generated characters, and automatically refunded if generation fails. For example, a 2,000-character audio drama episode (about 8 minutes in Chinese) consumes about 140 credits, and the ¥29.9 trial package is enough to make one episode. Role analysis can be used without registration; auditions and generation require purchasing credits. Credits do not expire and do not auto-renew.
Applicable scenarios include short drama going overseas, novel promotion posts, AI comic dramas, and audiobook production, and it also supports dubbing, multi-character dialogue, voice cloning, transcription subtitles, sound effects, and video translation dubbing, etc. In series works, each character can use the same voice in every chapter.
Limitations: a maximum of 8,000 characters can be processed at a time; auditions and generation require purchasing credits; specific prices and packages are subject to the official page, and other unmentioned functions, platform support, or effects are not currently provided.
