Models
Think of Fugu pricing as a single blended rate for the active agent pool, not a sum of every model used.
This is the “standard” Fugu model, recommended as the default. It aims for a balance between strong performance and low latency, making it ideal for daily use—such as coding, code reviews, chatbots, and interactive tasks. Depending on the specific task, Fugu routes the request to the most suitable model or agent. It also allows specific agents to be excluded from the pool to meet privacy, data, or compliance requirements.
It is the most powerful version. It coordinates a deeper pool of expert agents to maximize response quality for difficult or critical issues. It is designed for complex tasks such as Kaggle competitions, paper reproduction, cybersecurity analysis, and literature or patent research. It costs more and, according to the documentation, can route work among one to three agents depending on the problem.
agent is active エージェントが 1 つの場合 You pay only the standard rate for that specific underlying model.
With it, Fugu is charged at the standard rate of the underlying model, and when multiple agents are active we never stack fees: you pay a single rate based on the top tier model involved.
× the usage of Standard, built for heavy, long-running workloads.