YuriiFominYoung/sakana-fugu

YuriiFominYoung/ sakana-fugu:latest

109 Downloads Updated 1 week ago

Sakana Fugu is an advanced AI orchestration system developed by Tokyo-based Sakana AI.

cloud

ollama run YuriiFominYoung/sakana-fugu

curl http://localhost:11434/api/chat \
  -d '{
    "model": "YuriiFominYoung/sakana-fugu",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='YuriiFominYoung/sakana-fugu',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'YuriiFominYoung/sakana-fugu',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 1 week ago

1 week ago

3b2b761da3fa · 1.6kB ·

template

{{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}{{ if .Prompt }}<|im_start|>user

186B

system

You are the central coordinator of the Sakana Fugu framework, powered by a high-capacity, long-conte

897B

params

{ "num_ctx": 32768, "temperature": 0.3, "top_p": 0.9 }

48B

Readme

What is actually Sakana Fugu?

A Multi-Agent System, Delivered as One Model Sakana Fugu achieves superior performance by dynamically coordinating and orchestrating a diverse pool of powerful models. Instead of using domain knowledge to prescribe team organization, roles, or workflows, Fugu learns to dynamically assemble agents from a pool and coordinate them through non-obvious but highly efficient collaboration patterns.

Research-Driven Coordination for Multi-Agent Intelligence Sakana Fugu is grounded in two ICLR 2026 papers on learned model orchestration: TRINITY and the Conductor. Together, they show how systems can learn to assemble, route, and coordinate expert agents for each task instead of relying on hand-designed workflows. For a deeper look at the ideas behind the system, explore our technical report .

Sakana Fugu comes in two models — Fugu and Fugu Ultra .

Quantitative Results Our Fugu models surpass publicly accessible frontier models and are shoulder-to-shoulder with Fable 5 and Mythos Preview in various rigorous engineering, scientific, and reasoning benchmarks while delivering frontier capability without the risk of export controls.

Performance comparison of Fugu models and baseline frontier models across a suite of coding, reasoning, scientific, and agentic benchmarks. For Fable 5 and Mythos Preview, we report the max of the two if both scores are available on the same benchmark. Neither of them is in Fugu’s agent pool as they are not publicly accessible.

This model was made by YuriiFominYoung

What is actually Sakana Fugu?

A Multi-Agent System, Delivered as One Model
Sakana Fugu achieves superior performance by dynamically coordinating and orchestrating a diverse pool of powerful models. Instead of using domain knowledge to prescribe team organization, roles, or workflows, Fugu learns to dynamically assemble agents from a pool and coordinate them through non-obvious but highly efficient collaboration patterns.![fugu_arch_image.png](/assets/YuriiFominYoung/sakana-fugu/31580451-4c06-45ba-b290-5261fe804772)

Research-Driven Coordination for Multi-Agent Intelligence
Sakana Fugu is grounded in two ICLR 2026 papers on learned model orchestration: TRINITY and the Conductor. Together, they show how systems can learn to assemble, route, and coordinate expert agents for each task instead of relying on hand-designed workflows. For a deeper look at the ideas behind the system, explore our [technical report](https://arxiv.org/abs/2606.21228) .

Sakana Fugu comes in two models — Fugu and Fugu Ultra .

Quantitative Results
Our Fugu models surpass publicly accessible frontier models and are shoulder-to-shoulder with Fable 5 and Mythos Preview in various rigorous engineering, scientific, and reasoning benchmarks while delivering frontier capability without the risk of export controls.

![benchmark-chart.png](/assets/YuriiFominYoung/sakana-fugu/4f4fe8e1-66e5-4487-b180-dc756e9ef16f)

Performance comparison of Fugu models and baseline frontier models across a suite of coding, reasoning, scientific, and agentic benchmarks. For Fable 5 and Mythos Preview, we report the max of the two if both scores are available on the same benchmark. Neither of them is in Fugu’s agent pool as they are not publicly accessible.

This model was made by YuriiFominYoung

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)