Live LLM rankings based on benchmarks and real data from millions of people using models through OpenRouter.
Weekly usage of models across OpenRouter
Each task’s leading models, ranked by share of spend on OpenRouter
What one coding-agent session typically costs, by session length