Efficient AI text-to-speech model with 82M parameters delivering natural-sounding voice synthesis
Plan, edit and verify images with an AI agent that controls the workflow
Stage
Not given
Startup
Founded
Not given
2026
Based in
SG
Not given
Pricing model
Open source
Freemium
Pricing
Free to use. Free Chrome Extension with 10,000 characters free every month.
Free plan with daily tool allowances and 5 free agent runs. Creator plan $9.99/month with 700 monthly credits. Pro plan $29.99/month with 2,400 monthly credits. Studio plan $79.99/month with 7,000 monthly credits. Credit packs of 1,000 credits available for $15 and never expire. Light edits cost 10
Key features
82M parameter efficiency for lightweight, scalable performance
Multilingual support including English, French, Korean, Japanese, and Mandarin
Customizable voice packs with multiple lifelike voice options
Automatic content segmentation for chapter and section detection
OpenAI-compatible API endpoint for easy integration
Real-time audio generation with NVIDIA GPU acceleration
Processes up to 510 tokens in a single pass for longer audio outputs
Chrome extension for reading webpages aloud
Multi-step workflow planning with agent confirmation before execution
Protected details preserved during edits
Automatic image model selection per operation
Image to Prompt reverse-engineering in multiple output formats
Background removal with checkerboard preview
Image upscaling up to 4x with separate models for photos and line art
Image Prompt Generator with five variations exploring different axes
One credit pool shared across all tools
GPT Image 2.5 Flare and Sunburst model integration
Result verification comparing output against original request
What makes it different
Outperforms larger models like XTTS (467M params) and MetaVoice (1.2B params) despite having only 82M parameters
Open-source and free for both commercial and personal use under Apache 2.0 license
Lightweight architecture enabling deployment on CPU and GPU with minimal resources
Trained on high-quality, permissively licensed audio for accurate natural-sounding speech
Designed for scalability while preserving excellent audio quality
Plans and executes multi-step image workflows before running, unlike single-model generators
Protects specified image details that must remain unchanged during editing
Transparent pricing shown before any run; failed runs automatically refunded
Automatic model routing per operation type for optimal results