AivisSpeech alternative: when you need more than Japanese
AivisSpeech is a free, open-source Japanese text-to-speech app. It installs voice models from AivisHub, runs offline after the first launch and keeps an API compatible with VOICEVOX, so many tools that speak VOICEVOX also work with it.
This guide sums up what AivisSpeech offers, what to check before using it commercially, and when SuperClonica, a free engine with voices in 31 languages, is the better fit.
What AivisSpeech offers
According to the Aivis Project, AivisSpeech and its engine are LGPL-3.0 open source. They run on Windows and macOS, need the internet only on the first launch to download model data, and work on an ordinary CPU. Voice models come from AivisHub, and the UI and HTTP API are compatible with VOICEVOX.
Check the model licence
The software itself may be used free for personal and commercial work without credit, but each voice model keeps its own licence. AivisHub models are often published under ACML, ACML-NC or CC0, and an NC licence rules out commercial use, so read the model page before you publish.
When AivisSpeech is the better choice
For Japanese narration and a tool that plugs into VOICEVOX-based workflows, AivisSpeech is an excellent free option. SuperClonica’s catalog voices were mostly made from English and Russian recordings, so in Japanese they keep a trace of that accent.
When SuperClonica fits better
Choose SuperClonica when you need other languages: every voice speaks 31, from English and Spanish to Korean and Japanese. The catalog has narrators and game, anime and cartoon characters for $2.99 once or free, the ten stock voices are free under OpenRAIL-M, and the studio trains a voice from your own recording. Try any voice free on its page with 30 to 100 characters, or install the engine with python -m pip install superclonica-cpu and run superclonica --text-file script.txt --voice F2 --lang en --output narration.wav.