Fugu Ultra

Sakana AIMCP Tool CallReasoning

sakana/fugu-ultraClosed API

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

Benchmark Pricing$5.00 / 1M in$30.00 / 1M out$0.50 cache read

Context: 1,000,000 tokens•Max Output: 4,096 tokens•Released: 2026-06-15

TECHNICAL SPECIFICATIONS

Context Window1,000,0001M tokens

Max Output4,0964.1K tokens

LicenseProprietaryCloud API only

Input Modalities

TextVision

Function / Tool CallingSupported (MCP Ready)

Structured OutputsGuaranteed Schema

MODEL CONTEXT PROTOCOL (MCP) READINESS

✓ Fugu Ultra natively supports OpenAI / Anthropic format tool calling and is fully compatible with MCP servers and agent loops.

MCP Context Window Recommendations

With a context capacity of 1,000,000 tokens, this model can handle extremely large MCP resource manifests, complete database schemas, and multiple active server tool definitions simultaneously without degradation.

STANDARDIZED BENCHMARKS & EVALS

SWE Bench ProUnit not reported

73.7

Terminal Bench 2.1Unit not reported

82.1

LiveCodeBenchUnit not reported

93.2

LiveCodeBench ProUnit not reported

90.8

Humanity’s Last ExamUnit not reported

CharXiv ReasoningUnit not reported

86.6

GPQA DiamondUnit not reported

95.5

SciCodeUnit not reported

58.7

τ3 BankingUnit not reported

20.6

Long Context ReasoningUnit not reported

73.3

MRCRv2Unit not reported

93.6

CTI-REALMUnit not reported

69.4

API CALL INVOCATION EXAMPLE

// Example: Calling Fugu Ultra via OpenAI SDK
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.SAKANA_API_KEY,
});

const response = await client.chat.completions.create({
  model: "sakana/fugu-ultra",
  messages: [{ role: "user", content: "Hello! Can you inspect MCP tools?" }],
});

console.log(response.choices[0].message.content);