Back to News
Miso Labs•June 3, 2026
Miso TTS 8B
Open Source
text-to-speech
open-source
emotive speech
voice cloning
### TL;DR
Miso TTS 8B is an advanced text-to-speech model developed by Miso Labs, designed to generate high-quality, emotive speech from text input. Utilizing a hierarchical RVQ Transformer architecture, it offers state-of-the-art performance in conversational speech generation.
Key Insights & Metrics
Pricing
Free
Cost structure
Version
1.0
Current release version
Hardware
High-VRAM GPU; recommended VRAM: 24 GB for `bfloat16` precision, 40 GB+ for `float32` precision
Compute requirements
Category
Open Source
Licensing model
Region
United States
Primary region
Key Features
- 8-billion-parameter model
- Hierarchical RVQ Transformer architecture
- High-quality, emotive speech generation
- Open-source weights available on Hugging Face
- Supports English language only
- Intended for voice cloning, voiceover work, and educational content
Discussion
0
Upvotes
0
Downvotes
0 reviews
Sign in to leave a review
Reviews
No reviews yet. Be the first to review!