Source
IndexTTS2 official model/project documentation
Read the primary source.
Reference illustration, not a screenshot of this source.
IndexTTS 2 / 2.5
Speech-generation engine · Voice
Controllable zero-shot speech-generation system.
Use it for
Evaluate when precise control over the generated speech is needed.
Choose something else for
Do not treat a model listing as installed and ready.
Give it
Text; Reference audio where supported
Get back
Speech audio
How it is operated
Model runtime or compatible application
Availability
Provider capabilities and historical Studio notes are distinguished below. Current account access, credits and runtime readiness were not tested.
How charging works
Model compute; check the selected release and weights.
Where processing happens
Self-hosted/runtime dependent.
Language support
IndexTTS 2.5 documents Chinese, English, Japanese, Spanish and Arabic; do not assume French support.
Choose an underlying speech engine
★★★☆☆ 3/5 for this task
Editorial task fit based on the documented role and integration requirements. Check the released model and supported controls. A documented feature is not a verified local integration.
Decision: Evaluate when precise control over the generated speech is needed.
Limits and things to check
- Do not transfer capabilities between versions: the 2.0 release notes say precise duration control was not enabled; 2.5 documents a speaking-speed duration factor. Check supported languages.
Do not transfer capabilities between versions: the 2.0 release notes say precise duration control was not enabled; 2.5 documents a speaking-speed duration factor. Check supported languages.
Evidence
Official model/project documentation reviewed; no same-script listening test
Catalogue review: 2026-09-05. The evidence note identifies what was verified. This date does not imply a fresh tool test or verification of every linked source.
