Fugu Max
Neural Network
Fugu Max is a Sakana AI language model with multi-agent orchestration and request routing, available via Fugu Max API.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 000 000 tokens
- Max answer length
- 128 000 tokens
- Reasoning effort
- high, extra high, max
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Fugu Max?
Fugu Max is a model from the Fugu family by Sakana AI, focusing on the balance between quality and cost. It is not a monolithic neural network: Fugu is designed as a trainable multi-agent system where the language model parses the request and routes it along an appropriate path within the orchestration, and the response is assembled from coordinated steps. In BotHub, the model connects without a VPN or foreign card; you can pay with Russian cards in rubles based on actual usage, and unused tokens do not expire. Over 250 models are available in one window, making it easy to compare answers. A unified OpenAI-compatible API allows you to replace the model in a multi-model system without rewriting the integration. Chats are encrypted via AES-GCM, data is not saved, and companies have access to contracts, invoices, electronic document management (EDO), and an admin panel with limits. It is useful when you need to break down a complex task into steps, get a structured response, draft a document or email, analyze a technical description to create a summary, or run the same request through multiple models to choose the best option for production.Frequently asked questions about Fugu Max
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
fugu-max is a text model from Sakana with a 1,000,000 token context. It accepts not only text but also images and documents, can reason before answering, and call functions. Suitable for parsing large documents, working with code, analytics, and integrations.
In a single response, fugu-max generates up to 128,000 tokens — the volume of a small book. This is enough for long technical text, detailed code analysis, or translating a large document entirely without splitting the task. The input context is up to 1,000,000 tokens.
Yes, reasoning before answering is confirmed for fugu-max: the model breaks down the task into steps and checks itself before providing a result. This helps with math, STEM tasks, code debugging, and multi-step scenarios where accuracy is more important than speed.
Yes, fugu-max supports function calling and structured JSON output. The model is available via BotHub's unified OpenAI-compatible API, so it can be connected to agents, search, and internal services, and later switched to another model without rewriting the integration.
fugu-max accepts text, images, and documents as input, with documents processed natively without prior recognition. This makes it convenient to analyze screenshots, diagrams, tables, and multi-page files. There is no mention of audio or video in our data, so please rely on the confirmed formats.
Our data does not describe the quality of specific languages, so we cannot make promises — please test the model with your task. The BotHub interface and payment are in rubles, access from the Russian Federation is possible without a VPN or foreign card, and you can compare it with another model in the same window.