No model loaded

Select Model

Choose an AI model based on your device capabilities

Detecting hardware capabilities...

TinyLlama 1.1B

Smallest model, runs on any device including WASM fallback

Size:0.6GB
Tiers:MINIMAL, LOW, MID, HIGH

Gemma 2 2B Instruct

Balanced performance and quality for mid-range devices

Size:1.4GB
Tiers:LOW, MID, HIGH

Llama 3.2 3B Instruct

High-quality responses for capable devices

Size:1.9GB
Tiers:MID, HIGH

Phi-3.5 Mini Instruct

Best quality, requires high-end hardware

Size:2.1GB
Tiers:HIGH