Fugu Max: What It Does and What It Costs
sakana/fugu-maxFugu Max is Sakana’s text-and-image model, built to coordinate multiple AI agents.
1M tokens of context — about 750,000 words, on a scale that ends at 2M.
What it can do
- Reasoning
- Vision
- Tool use
- Takes text
- Takes image
- Takes file
- Gives text
Released 11 Sept 2026 · checked 3 Oct 2026.
About it
Fugu Max is a Sakana model that uses a learned system to route work among multiple AI agents rather than relying on one model alone. It accepts text, images, and files, and can use tools. It has a one-million-token context window and costs $2 per million input tokens and $6 per million output tokens.
Review
In a line
Fugu Max is a Sakana model that coordinates multiple AI agents to handle a task. It takes text, images, and files as input and produces text.
What it's good at
Its reasoning and tool-use abilities suit tasks that benefit from working through information or using tools. It can also take images as input.
Price in plain terms
Fugu Max costs $2 per million input tokens and $6 per million output tokens. Input is what you send; output is what the model writes. A page of about 500 words costs about 0.13 cents to read. A chat turn with about 1,000 input tokens and 500 output tokens costs about 0.5 cents; 1,000 such turns cost about $5.00.
Context and limits
Fugu Max can take in up to one million tokens at once, so it can work with a large amount of text and other input. It can write up to 128,000 tokens in one answer. Its output is text, though it can accept images and files as input.
Ways to use it
Fugu Max is a model from Sakana. The facts describe its inputs, tools, and pricing, but do not specify apps or providers through which people can access it.
Verdict
Consider Fugu Max if you need a model that can reason, use tools, and take text, images, or files as input, especially when a large context window matters. Its per-token prices and sample costs make it possible to estimate usage.
Verdict
Fugu Max may fit people handling large volumes of material who want a model that can reason, use tools, and accept images or files alongside text. Its million-token context window gives it room for extensive input, while the sample costs offer a basis for estimating text-chat use. Look elsewhere if you need the model to create images, or if you need a confirmed connection to a specific app or provider; those details are not established here.
- Accepts text, images, and files as input.
- Can reason and use tools.
- One-million-token context window.
- Sample costs are provided for pages and chat turns.
- Produces text only.
- The facts do not name apps or access providers.
- A single chat turn costs more for output tokens than input tokens.
- Pick it ifPeople working with large amounts of text and other input; Tasks that benefit from reasoning or tool use
- Skip it ifYou need the model to generate images.; You need a confirmed app integration or provider access route.
Pricing
Fugu Max costs $2 per million input tokens and $6 per million output tokens. A page of about 500 words costs about 0.13 cents to read; a sample chat turn costs about 0.5 cents, or about $5.00 for 1,000 turns.
Questions
Who makes Fugu Max?
Sakana makes Fugu Max.
What can I give Fugu Max?
It accepts text, images, and files as input.
Can Fugu Max use tools?
Yes. It supports tool use and reasoning.
How much can it read at once?
Its context window is one million tokens.
How long can its answer be?
Its maximum output is 128,000 tokens.
How much does a chat turn cost?
A turn with about 1,000 input tokens and 500 output tokens costs about 0.5 cents. 1,000 such turns cost about $5.00.
Does Fugu Max make images?
No image output is listed; the model outputs text, though it can accept images as input.