Model Comparison

Performance, cost, and reliability across all active models

Live · refreshing in 30s

Select exactly 2 models to see comparison diff

Cost by Model

Total USD spend per model in period

Latency Comparison

Avg and P95 response time per model

P95 Latency Distribution

Average vs tail latency per model — high spread means unpredictable responses

Multi-Dimension Comparison

Normalized 0–100 score across 5 dimensions. Higher = better in all axes.