Your Device vs SOTA AI

Detected Hardware

🖥️

Detecting...

Device Type

🧠

RAM (GB)

CPU Cores

🎮

Detecting...

GPU Tier

Your AI Hosting Capability

Calculating...

Analyzing your hardware configuration to determine the most capable model you can run locally with good performance...

You

Entry (1B)<br>Mid (7B)<br>Large (70B)<br>SOTA (1T+)

—%

away from top-tier SOTA

Model Tier Comparison

Model Tier<br>Example Models<br>Params<br>RAM Needed<br>Your Status

Note: These estimates assume quantized models (4-bit) for consumer hardware. Your actual performance depends on quantization quality, model architecture, and optimization tools (llama.cpp, vLLM, etc.). SOTA models like GPT-4o, Claude 3.5, and Gemini Ultra are estimated at 1T+ parameters.

🔄 Re-detect Hardware