peculiar.cloud AI

Updated daily · OpenRouter-compatible

AI Model Rankings

A capability score from Artificial Analysis quality and LM Arena performance — price is handled by the tiers, not the score, so a model ranks on how good it is. Compare text and vision models by score, cost, context window, and reasoning support, then use the model IDs directly with OpenRouter.

Capability score (0–100): Artificial Analysis quality 60% + LM Arena ELO 40%. Price is not a factor — the tiers handle cost. n/2 cross-source matches — how many of Artificial Analysis and LM Arena the model was matched against.
Loading...