atodekiku-supertonic-3

AtodeKiku Supertonic 3 mirror

This repository is an independently maintained, byte-identical distribution mirror for AtodeKiku.

It is not the official Supertone repository and does not imply endorsement by Supertone Inc.

The distribution repository is umiyosh/atodekiku-supertonic-3.

Provenance

• Upstream model: Supertone/supertonic-3
• Upstream model revision: 3cadd1ee6394adea1bd021217a0e650ede09a323
• Upstream inference source commit: 7e2804f96016a7028cb1ed627353c61c1e9dd281
• Mirror artifact count: 16
• Mirror artifact bytes: 401,276,744
• Artifact modifications: none

The files listed in manifest.json are byte-identical to the upstream model revision.

AtodeKiku verifies the byte size and SHA-256 of every artifact before installation.

Scope

This mirror contains the four ONNX models, TTS configuration, Unicode indexer, and ten preset voice styles required by AtodeKiku.

Sample audio and promotional images are not part of AtodeKiku's runtime distribution contract.

License and use restrictions

The model artifacts are distributed under the BigScience OpenRAIL-M License in LICENSE.

Use and redistribution remain subject to Section 5 and Attachment A of that license.

All original copyright, patent, trademark, and attribution notices remain in effect.

The inference sample source is separately licensed under the MIT License in the upstream source repository.

Copyright (c) 2026 Supertone Inc.

Immutable downloads

Consumers must use a destination commit hash in Hugging Face Resolver URLs instead of the mutable main branch.

manifest.json records the path, byte size, and SHA-256 of each runtime artifact.