Total tool/function calls made through OpenRouter — indicates how much each model is used in agentic, tool-using workflows.
Highlights which models are actually deployed in function-calling and agent pipelines; complements benchmark-only views with observed production behavior.
Volume can reflect SDK defaults or integration maturity instead of underlying reasoning quality; tool-call counts do not indicate whether calls were successful or useful.
Higher is better for agentic adoption signal. High tool-call volume indicates frequent production use in function-calling workflows.
Volume can reflect SDK defaults or integration maturity instead of underlying reasoning quality.
No scores available for this benchmark yet.