Select Your Intelligence
Pick a specialized model to boot up on your hardware.
Gemini Nano
Recommended
Powerful on-device AI for private, local chat.
NATIVE-API
N/A
Qwen 3 0.6B
Recommended4GB RAM
Alibaba's latest ultralight model. Extremely fast and efficient.
WEB-LLM-Q4
~800MB
Qwen 2.5 0.5B
Recommended4GB RAM
Verified stable lightweight model. Best for basic devices.
WEB-LLM-Q4
~800MB
Gemma 2 2B
Recommended8GB RAM
Google's updated 2B model with superior reasoning over v1.
WEB-LLM-Q4
~2GB
Qwen 3 4B
Recommended8GB RAM
Powerful 4B parameter model for high-end local reasoning.
WEB-LLM-Q4
~2GB
Mistral 7B v0.3
12GB RAM
Mistral v0.3 with native function calling support.
WEB-LLM-Q4
~4GB+
Llama 3 8B
16GB RAM
Professional grade intelligence for complex writing tasks.
WEB-LLM-Q4
~4GB+
Initial launch may download significant data depending on selection. WebGPU or Chrome AI support required.