← Model list

Fugu

sakana·sakana/fugu·GA·Closed·fugu series·NEW

Multi-agent model for routing expert agents across complex analytical tasks

At a glance
Official pricePrice undisclosed
Lowest paidNo paid channels
Context1,000,000
Output limit1,000,000
Capabilities
ReasoningTool useStructured outputTemperatureAttachments
Modalities
TextImage
Knowledge cutoff
Released / updated2026-06-15 / 2026-06-15

Quality & performance

Artificial Analysis doesn't cover this model (267 of 2059 have data). Quality data comes from independent evals covering widely used models.

Available at 1 providers0 with public prices

ProviderTierInputOutputCache readCache writeContextOutput limitStatus
Sakana AIOfficialFirst-party1,000,0001,000,000hostTable.undisclosed

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

This model has no public prices; can't estimate.

Benchmark12 items

NameConditionsScoreMetricSource
SWE Bench Pro59Source ↗
Terminal Bench 2.180.2Source ↗
LiveCodeBench92.9Source ↗
LiveCodeBench Pro87.8Source ↗
Humanity’s Last Exam47.2Source ↗
CharXiv Reasoning85.1Source ↗
GPQA Diamond95.5Source ↗
SciCode60.1Source ↗
τ3 Banking21.7Source ↗
Long Context Reasoning74.7Source ↗
MRCRv286.6Source ↗
CTI-REALM67.5Source ↗

The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.

Reasoning control

effort = higheffort = xhigh

Related models

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$0.0001$02026-08-052026-08-13