sakana
Fugu Ultra v2
Fugu Ultra coordinates a deeper pool of expert agents to maximize answer quality on hard, high-stakes problems. Fugu Ultra focuses on maximizing performance, for a higher cost.
reasoningtool usevisionimplicit cachingstructured outputtextimage
Overview
- Type
- language
- Context
- 1M
- Max output
- 1M
- Input / output
- text, image / text
API
Model identifier
sakana/fugu-ultra-v2Providers and live performance
Rolling Gateway metrics. Latency is time to first token; throughput is output tokens per second.
| Provider | Status | Latency | Throughput | Uptime 15m / 1h / 1d | Input | Output | Cache read / write | Discount | Regions | ZDR | No training |
|---|---|---|---|---|---|---|---|---|---|---|---|
| sakana | Operational | 4.58s / 87.71s | — | 100.00% / 100.00% / 99.65% | $5.00/M | $30.00/M | $0.50/M / — | — | — | No | Yes |
Pricing
input
0.000005
input tiers
{
"cost": "0.000005",
"min": 0,
"max": 272001
}, {
"cost": "0.00001",
"min": 272001
}
output
0.00003
output tiers
{
"cost": "0.00003",
"min": 0,
"max": 272001
}, {
"cost": "0.000045",
"min": 272001
}
input cache read
0.0000005
input cache read tiers
{
"cost": "0.0000005",
"min": 0,
"max": 272001
}, {
"cost": "0.000001",
"min": 272001
}
Capabilities
reasoningtool usevisionimplicit cachingstructured outputtextimage
Supported parameters
max tokens
Yes
temperature
Yes
stop
Yes
tools
Yes
tool choice
Yes
reasoning
Yes
include reasoning
Yes
response format
Yes
structured outputs
Yes
Reasoning options
options
{
"type": "effort",
"values": [
"high",
"xhigh",
"max"
]
}
Metadata
type
language
context window
1M
max output
1M
released
Sep 10, 2026
temperature control
Yes
zero data retention
none
no training
all
api specifications
v2, v3, v4