KaniTTS

Overview

• Model Size: 450M parameters
• Sample Rate: 22kHz
• Languages: English, German, Arabic, Chinese, Korean, French, Japanese, Spanish
• License: Apache 2.0

Performance

• Latency: ~1 second to generate 15 seconds of audio
• Memory: 2GB GPU VRAM
• Quality Metrics: MOS 4.3/5 (naturalness), WER <5% (accuracy)

• Dataset: ~80k hours from LibriTTS, Common Voice, and Emilia
• Hardware: 8x H100 GPUs, 45 hours training time on Lambda AI

Audio Examples

Use Cases

• Conversational AI: Real-time speech for chatbots and virtual assistants
• Edge/Server Deployment: Resource-efficient inference on affordable hardware
• Accessibility: Screen readers and language learning applications
• Research: Fine-tuning for specific voices, accents, or emotions

Limitations

• Performance degrades with inputs exceeding 2000 tokens
• Limited expressivity without fine-tuning for specific emotions
• May inherit biases from training data in prosody or pronunciation
• Optimized primarily for English; other languages may require additional training

Optimization Tips

• Multilingual Performance: Continually pretrain on target language datasets and fine-tune NanoCodec
• Batch Processing: Use batches of 8-16 for high-throughput scenarios
• Hardware: Optimized for NVIDIA Blackwell architecture GPUs

Resources

• Pretrained Model
• Fine-tuned Model
• HuggingFace Space

• Inference Example
• Fine-tuning-code
• Example Dataset
• GitHub Repository

• Website
• Contact Form

Acknowledgments

Responsible Use

• Illegal content or harmful, threatening, defamatory, or obscene material
• Hate speech, harassment, or incitement of violence
• Generating false or misleading information
• Impersonating individuals without consent
• Malicious activities such as spamming, phishing, or fraud

Contact