Fugu Ultra

Tool callingStructured outputReasoningImage input

Website

Quality-first multi-agent model for hard research, analysis, and competitions

Context: 1M
Released: 4mo ago
Input / 1M tokens: $5.00
Output / 1M tokens: $30.00
Cached input / 1M tokens: $0.50

Sources:models.devOpenRouter

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

No benchmark results have been matched to this model yet.

Specifications

Developer: Sakana
Model family: Fugu
Type: Language model
Predecessor: —
Successor: —
Released: June 15, 2026
Model ID: sakana/fugu-ultra
Openness: Closed
License: —
Input modalities: text, image
Output modalities: text
Modality: text+image->text
Tokenizer: Other
Max output tokens: 128K
Hugging Face: —
Source freshness: Metadata synced Oct 8, 2026, 10:17 PM UTC

Additional source evidence

Source model: sakana/fugu-ultra
Source family: fugu
Last updated: 2026-06-15
Model context: 1M
Open weights: No
File attachments: Yes
Reasoning: Yes
Tool calling: Yes
Structured output: Yes
Temperature control: No

Published benchmarks

Benchmark	Score	Evaluation details	Evidence
SWE Bench Pro	73.7	—	Source
Terminal Bench 2.1	82.1	—	Source
LiveCodeBench	93.2	—	Source
LiveCodeBench Pro	90.8	—	Source
Humanity’s Last Exam	50	—	Source
CharXiv Reasoning	86.6	—	Source
GPQA Diamond	95.5	—	Source
SciCode	58.7	—	Source
τ3 Banking	20.6	—	Source
Long Context Reasoning	73.3	—	Source
MRCRv2	93.6	—	Source
CTI-REALM	69.4	—	Source

Official references

Official model catalog

Complete source record (JSON)

{
  "id": "sakana/fugu-ultra",
  "name": "Fugu Ultra",
  "description": "Quality-first multi-agent model for hard research, analysis, and competitions",
  "family": "fugu",
  "last_updated": "2026-06-15",
  "attachment": true,
  "reasoning": true,
  "tool_call": true,
  "structured_output": true,
  "temperature": false,
  "modalities": {
    "input": [
      "text",
      "image"
    ],
    "output": [
      "text"
    ]
  },
  "limit": {
    "context": 1000000
  },
  "benchmarks": [
    {
      "name": "SWE Bench Pro",
      "score": 73.7,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "Terminal Bench 2.1",
      "score": 82.1,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "LiveCodeBench",
      "score": 93.2,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "LiveCodeBench Pro",
      "score": 90.8,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "Humanity’s Last Exam",
      "score": 50,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "CharXiv Reasoning",
      "score": 86.6,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "GPQA Diamond",
      "score": 95.5,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "SciCode",
      "score": 58.7,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "τ3 Banking",
      "score": 20.6,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "Long Context Reasoning",
      "score": 73.3,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "MRCRv2",
      "score": 93.6,
      "source": "https://console.sakana.ai/models"
    },
    {
      "name": "CTI-REALM",
      "score": 69.4,
      "source": "https://console.sakana.ai/models"
    }
  ],
  "release_date": "2026-06-15",
  "open_weights": false,
  "links": [
    {
      "url": "https://raw.githubusercontent.com/SakanaAI/fugu/refs/heads/main/configs/files/fugu.json",
      "label": "Official model catalog",
      "type": "docs"
    }
  ]
}

Infrastructure

Provider availability

All providers

Provider	Endpoint	Context	Input /M tokens	Output /M tokens	Cached input /M tokens	Cache write /M tokens	Uptime 1d	Data handling
Sakana AI	sakana	1M	$5.00	$30.00	$0.50	—	100.00%	ZDR: No Training: Unknown Prompt logging: Unknown Retention: Unknown ZDR source Privacy policy Checked 2026-10-15

Provider data:OpenRouter