Fugu Ultra
sakana/fugu-ultra
Quality-first multi-agent model for hard research, analysis, and competitions
Capabilities
ReasoningYes
Tool callingYes
Structured outputYes
TemperatureFixed
AttachmentYes
Open weightsNo
Limits
Context1,000,000 tokens
Max inputN/A
Max outputN/A
Details
Familyfugu
Knowledge cutoffN/A
Release date2026-06-15
Last updated2026-06-15
Modalities
Inputtext, image
Outputtext
Benchmarks
SWE Bench Pro
73.7
Terminal Bench 2.1
82.1
LiveCodeBench
93.2
LiveCodeBench Pro
90.8
Humanity’s Last Exam
50
CharXiv Reasoning
86.6
GPQA Diamond
95.5
SciCode
58.7
τ3 Banking
20.6
Long Context Reasoning
73.3
MRCRv2
93.6
CTI-REALM
69.4
Community Rating
—
No ratings yet
Sign in to rate
Pricing
Cost per 1M tokens
InputN/A
OutputN/A