Fugu Ultra is a Sakana AI model tracked in Agent Serve. It supports a 1M token context window. Pricing starts at $5/1M input tokens and $30/1M output tokens. Key capabilities include Tool choice, Structured outputs. Best for long-context retrieval, large documents, and high-memory workflows.
Best forBest for long-context retrieval, large documents, and high-memory workflows.
TemperatureNot configurable
Reasoning effortNot supported
VerbosityNot supported
Thinking levelsNot supported
Structured outputsSupported
Tool choiceSupported
Computer useNot supported
Deep researchNot supported
Memory supportSupported
Max output tokensNot published
Frequently asked questions
Fugu Ultra is a Sakana AI model available in Agent Serve. Fugu Ultra is a Sakana AI model tracked in Agent Serve. It supports a 1M token context window. Pricing starts at $5/1M input tokens and $30/1M output tokens. Key capabilities include Tool choice, Structured outputs.
Fugu Ultra starts at $5/1M input tokens, $0.5/1M cached input tokens, and $30/1M output tokens. Above 272k input tokens, the full request is priced at $10/1M input tokens, $1/1M cached input tokens, and $45/1M output tokens.
Fugu Ultra supports a context window of 1M tokens in Agent Serve. In an agent, this determines how much conversation history, tool outputs, and retrieved documents the model can hold in a single call.
Fugu Ultra supports the following capabilities in Agent Serve: Tool choice, Structured outputs.
Best for long-context retrieval, large documents, and high-memory workflows. When used in an Agent Serve workflow, it can be selected in any Agent block from the model picker.
