YuriiFominYoung/sakana-fugu-ultra

YuriiFominYoung/ sakana-fugu-ultra:latest

172 Downloads Updated 1 week ago

Sakana Fugu Ultra is the flagship, performance-oriented tier of the "Fugu" multi-agent AI system developed by Tokyo-based startup Sakana AI.

cloud

ollama run YuriiFominYoung/sakana-fugu-ultra

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

from ollama import chat

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

import ollama from 'ollama'

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

Details

Updated 1 week ago

1 week ago

d2b75407bc6f · 1.8kB ·

system

You are the elite orchestration intelligence of the Sakana Fugu Ultra framework. Your architecture i

1.1kB

params

{ "num_ctx": 65536, "temperature": 0.25, "top_p": 0.9 }

49B

template

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

186B

Readme

Fugu Ultra coordinates a deeper pool of expert agents to maximize answer quality on hard, high-stakes problems. Early users rely on it for Kaggle competitions, paper reproduction, cybersecurity analysis, and literature and patent investigations.

Quantitative Results

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