Intelligence: Measures reasoning quality, logic depth, and decision consistency. Score range 0–10; higher scores indicate stronger reasoning performance.

Capacity: Measures how much context the model can handle coherently. Score range 0–10; higher scores indicate better long-context capability.

Experience: Measures response speed, interaction smoothness, and output stability. Score range 0–10; higher scores indicate better user experience.

Trend: Measures multimodal strength and adaptability to future features. Score range 0–10; higher scores indicate stronger trend competitiveness.

Economy: Measures cost-efficiency balancing quality, speed, and price. Score range 0–10; higher scores indicate better value for real usage.

Trust: Measures reliability, safety control, and output risk resilience. Score range 0–10; higher scores indicate stronger trustworthiness.