One API key for accessing 268+ AI models across 45+ providers with automatic failover and smart routing
Foundation model for mobile, wearables, robots, smart home, automotive and microcontrollers
Stage
Not given
Startup
Founded
Not given
Not given
Based in
Not given
Not given
Pricing model
Freemium
Open source
Pricing
Free: 141 free models with no cost. Pay as you go: add credit from $1 to $10,000 with no expiration. Subscriptions: Starter $20/month for $40 API credit, Pro $50/month for $100 API credit, Scale $100/month for $200 API credit. All plans provide 2x credit multiplier on subscriptions.
Not given
Key features
OpenAI-compatible API endpoint
Access to 268+ AI models from 45+ providers
Automatic provider failover
Support for Claude, GPT, Gemini, and DeepSeek models
Built-in character chat client
Support for SillyTavern, RisuAI, and character chat clients
Works with coding agents like Claude Code, Cline, and Roo Code
Multi-protocol support including Anthropic and Gemini compatibility
Support for up to 1M token context windows
Private conversations stored in browser only
Tool calling with function selection and argument filling
Structured extraction with guaranteed JSON parsing
Text embeddings for local semantic search and matching
Confidence scoring for routing and gating calls
Laddered subnetwork architecture from 2 to 20 layers
LoRA fine-tuning support for customization
400-4k tokens per second inference speed on Raspberry Pi 5
Python package with pip install and Hugging Face distribution
What makes it different
One API key for all models instead of managing multiple provider keys
Automatic provider failover with transparent retry logic
No model gating by subscription tier
Built-in chat client with character support and no server storage
Pay-as-you-go with no expiration or subscription required
Beats models 10x its size on mobile tool calling benchmarks
Matches models 2-3x larger on structured extraction tasks
Entire model in a single 8-29 MB binary for tiny devices
Simple Attention Network architecture uses 2x fewer MFLOPs than transformers
Fine-tuned subnetworks outperform larger models like DeepSeek V4 Flash
UnoRouter