Fugu Ultra v1.1
Public AI IQ model profile with source-backed benchmark results and derived ranking context.
IQ
118
EQ
n/a
Effective Cost
$37.01
IQ Rank
#34
Fugu Ultra v1.1 API pricing — full price table, workload cost examples, and cheaper alternatives.
IQ Dimensions
| Dimension | IQ | Coverage |
|---|---|---|
| Abstract Reasoning | 101 | 0/3 |
| Mathematical Reasoning | 118 | 0/5 |
| Academic Reasoning | 130 | 1/6 |
| Programmatic Reasoning | 123 | 1/6 |
| Computer Use | 126 | 1/6 |
| Reliability | 112 | 0/7 |
Benchmark Results
| Benchmark | Score | Dimension |
|---|---|---|
| GPQA Diamond | 95.6 | Academic Reasoning |
| Terminal-Bench 2.1 | 90 | Programmatic Reasoning |
| Toolathlon | 75 | Computer Use |
Compare
Frequently Asked Questions
- What should I know about Fugu Ultra v1.1?
- Sakana AI develops this multi-model orchestration system in Japan. Results measure the coordinated model pool, not a standalone base model. These are provider-reported system evaluations, not independent Artificial Analysis or Vals measurements. Benchmark effort and harness details are not fully disclosed in the newer charts. The IQ estimate has 3 source-backed benchmarks across 3 dimensions. Missing coverage may use scoring-only estimates; predecessor values are not new measurements. Sakana reports HLE (text), but its tool configuration and equivalence to the tracked HLE protocol are unverified. ProgramBench metric and DeepSWE version are also unverified; those values are not mapped into existing IQ or domain fields. Standard pricing is $5 input and $30 output per million tokens; above 272K context it is $10/$45. Orchestration tokens are billable. No measured orchestration usage is available for effective-cost estimation. Missing benchmark estimates follow the explicit Ultra predecessor chain. The versioned endpoint is distinct from the mutable fugu-ultra alias. The 1M-token context window and text/image inputs are documented in Sakana’s client model catalog; context length is not inferred from pricing thresholds.
- What is the IQ of Fugu Ultra v1.1?
- Fugu Ultra v1.1 has an estimated IQ of 118, which ranks #34 out of 140 public models on the AI IQ leaderboard. The score combines six equally weighted reasoning dimensions, each measured from source-backed benchmark results.
- How much does Fugu Ultra v1.1 cost?
- Fugu Ultra v1.1 costs $5.00 per million input tokens and $30.00 per million output tokens. Adjusted for how many tokens it actually consumes on real tasks, its effective cost works out to $37.01 per million input-plus-output tokens.
- Is Fugu Ultra v1.1 open source?
- No — Fugu Ultra v1.1 is a proprietary model. Its weights are not publicly available, so it is used through hosted APIs and products.
- When was Fugu Ultra v1.1 released?
- Fugu Ultra v1.1 was publicly released on July 24, 2026.
- How large is the context window of Fugu Ultra v1.1?
- Fugu Ultra v1.1 supports a context window of 1,000,000 tokens — the amount of text it can consider in a single request.