Skip to content

SOTA-Class TTS at Compact Scale

Python
5.5 KB
Raw Blame

Audio8-TTS-Preview-0.6B

All releases · 中文 · Model · Demo

This page preserves the original Preview 0.6B language coverage, architecture, and evaluation. Realistic Air's 23-language coverage and new results belong to that release; Preview keeps its own model and usage instructions.

Quick start

Follow the installation instructions, then run from the repository root:

python audio8_tts_infer.py \
  --model Edge0/Audio8-TTS-Preview-0.6b \
  --text "Welcome to Audio8 TTS Preview." \
  --output outputs/preview.wav

For cloning, batching, and training, continue to select Edge0/Audio8-TTS-Preview-0.6b.

Inference · SGLang Omni · ONNX INT4 · SFT

Supported languages

The Preview checkpoint performs best in the following languages:

Language Name
Cantonese 粤语
Chinese 中文
Dutch 荷兰语
English 英语
French 法语
German 德语
Italian 意大利语
Japanese 日语
Korean 韩语
Polish 波兰语
Spanish 西班牙语

Architecture

audio8_tts uses a DualAR architecture inspired by Fish Audio S2 Pro.

Component Configuration
Main model 601,159,424 parameters, excluding the codec
Slow AR 24 layers, width 896, 14 attention heads, 2 KV heads
Fast AR 4 layers, width 896, 14 attention heads, 2 KV heads
Acoustic tokens 10 codebooks, 4,096 entries per codebook
Codec 44.1 kHz, 2,048 samples per model frame (~21.5 frames/s)
Context Up to 2,048 packed text/audio positions

The slow AR transformer predicts one semantic token for each audio frame. The fast AR transformer then predicts the frame's codec codebooks, conditioned on the slow hidden state and preceding codebooks. Static KV caches are used by both branches during generation. The checkpoint also bundles its neural codec, so reference encoding and waveform decoding require no separate model.

Evaluation

Audio8 TTS Preview is the smallest model in this comparison at just 0.6B parameters. Despite using only a fraction of the parameters of the other systems, it delivers results in the first tier of industry-leading SOTA TTS models on the benchmarks below. In particular, it achieves the best English WER and competitive Chinese CER on Seed-TTS, while remaining competitive across the CV3 multilingual evaluation.

Lower WER/CER is better; higher SIM is better. Seed-TTS similarity values are shown as percentages.

Seed-TTS

Model Parameters EN WER / SIM ZH CER / SIM Hard ZH CER / SIM
Audio8 TTS Preview 0.6B 1.506 / 63.2 0.950 / 73.1 11.510 / 68.7
Fish S2 Pro 4.6B 1.607 / 64.6 1.038 / 73.8 10.149 / 70.1
Higgs Audio v2 4.7B 1.524 / 66.4 0.806 / 72.1 10.622 / 69.3
CosyVoice3-1.5B 1.5B 2.22 / 72.0 1.12 / 78.1 5.83 / 75.8
MOSS-TTS 8.5B 1.85 / 73.4 1.20 / 78.8 -
VoxCPM2 2.3B 1.84 / 75.3 0.97 / 79.5 8.13 / 75.3

Seed-TTS WER and CER comparison

CV3 multilingual error rate

Model Parameters zh en hard-zh hard-en ja ko de es fr it ru
Audio8 TTS Preview 0.6B 3.205 3.128 10.535 5.997 7.205 4.223 3.447 3.641 8.790 4.790 -
Fish S2 Pro 4.6B 3.600 3.493 10.588 7.349 5.139 4.111 3.605 2.972 8.600 4.229 4.702
Higgs Audio v2 4.7B 3.378 3.404 10.424 5.754 4.742 4.260 3.300 2.929 9.425 3.555 5.423
CosyVoice3-1.5B 1.5B 3.91 4.99 9.77 10.55 7.57 5.69 6.43 4.47 11.8 10.5 6.64
VoxCPM2 2.3B 3.65 5.00 8.55 8.48 5.96 5.69 4.77 3.80 9.85 4.25 5.21

CV3 multilingual WER and CER comparison

Parameter counts are calculated directly from the released weight tensors. MOSS-TTS contains 8,489,841,664 parameters. VoxCPM2's main model contains 2,290,004,544 parameters; the separate AudioVAE is not included in the parameter comparison.

Fish S2 Pro was reevaluated because its official evaluation uses its own normalizer. Higgs Audio v2 was evaluated locally because concrete values were unavailable. All other baseline values were collected from their official reports through the VoxCPM repository.

Different normalizers and evaluators make cross-project values reference comparisons rather than a strictly matched ranking. Evaluation coverage does not expand the Preview's supported-language claim beyond the 11 languages listed above.

Limitations and responsible use

  • This is a Preview checkpoint with limited multilingual and dialect coverage.
  • Very long, noisy, or inaccurate reference clips can reduce stability and speaker similarity.
  • Generated speech can be misused for impersonation or misinformation. Obtain consent before cloning a voice and clearly disclose synthetic audio where appropriate.
  • Test the model for accuracy, safety, and legal compliance before deployment.

License

The code and these 0.6B weights use Apache License 2.0. NOTICE.