Fugu Ultra

Fugu Ultra is a multi-agent model from Sakana AI, designed for hard research, analysis, and competitions. It is part of the fugu line, succeeding the earlier Fugu model. It has a context window of 1,000,000 tokens and a maximum output of 1,000,000 tokens. Pricing starts at $5.00 per million input tokens and $30.00 per million output tokens, with cached input at $0.50 per million. The weights are not open. It supports reasoning mode, tool calling, and accepts image and text inputs, returning text.

First version of its line in the registry.

• ·1M token context window
• ·1M token max output
• ·Multi-agent architecture for complex tasks
• ·Supports reasoning and tool calling
• ·Accepts image and text inputs

• Reach for it for hard research tasks
• Reach for it for in-depth analysis
• Reach for it for competitive problem solving
• Reach for it for long-context document processing

11 hosts serve this model at the price above · the maker's documentation

The registry keeps the six most recent versions of a line. Older ones stay in the file and can be surfaced without a rebuild.

What is the context window size?

The context window is 1,000,000 tokens, and the maximum output is also 1,000,000 tokens.

Is Fugu Ultra open source?

No, the weights are not open. It is served by 11 hosts.

What are the pricing details?

Input costs from $5.00 per million tokens, output $30.00 per million, and cached input $0.50 per million.