NOUMENOpen app
NOUMEN

sakana · Frontier

Fugu Ultra v1.1

Added Jul 24, 2026

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Sakana AI's upgraded Fugu Ultra release with stronger coding, agentic task execution, and advanced reasoning through dynamic orchestration of frontier models.

  • Reasoning
  • Tools
  • Vision
  • Structured output
Input
Text, Images
Output
Text
Reasoning effort
high · xhigh · max
Discovery tier
Frontier

Pricing

USD per 1M tokens
Input
$5.00
Output
$30.00
Cache read
$0.50

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
16.4K

Side by side

Fugu Ultra v1.1 compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Fugu Ultra v1.1sakanaThis modelClaude Opus 5AnthropicRemove
ProvidersakanaAnthropic
Discovery tierFrontierFrontier
AddedJul 24, 2026Jul 24, 2026
Limits
Context window1M1M
Max output16.38K128KBest
Average output—1.5K tokens
Pricing, USD per 1M tokens
Input$5.00$5.00
Output$30.00$25.00Best
Cache read$0.50$0.50
Cache write—$6.25
Performance
Throughput—177.8 tok/s
Time to first token—3.8 s
Request success (24h)—99.9%
Benchmark indexes
Intelligence Index—48.1
Coding Index—76.5
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesYes
PDF inputNoYes
Reasoning efforthigh, xhigh, maxlow, medium, high, xhigh, max

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • Fugu Ultra v1.1
  • Claude Opus 5

Intelligence Index

  • Fugu Ultra v1.1—
  • Claude Opus 548.1
3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 76.5

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 53.6%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 25.3%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 1,558 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 1,595 Elo

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 19.6%

  • AA-LCRv1.1

    Reasoning across long documents

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 79.0%

  • MLCR-AA

    Long-context reasoning over medical records

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 11.7%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 52.8%

  • GPQA Diamond

    Graduate-level science questions

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 93.7%

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 46.0%

  • SciCode

    Scientific computing in Python

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 55.4%

Legacy evaluations

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Fugu Ultra v1.1: —
    • Claude Opus 5: 79.0%

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "sakana/fugu-ultra-v1.1",
    "reasoning_effort": "high",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="sakana/fugu-ultra-v1.1",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

sakana

Other sakana models, starting with the same family.