Benchmark scores
Higher is better for all benchmark scores (SWE-bench, GPQA, HLE, Terminal-Bench, etc.). Green marks the model with the higher score.
LLM Comparison
Sakana Fugu Ultra vs Gemini 3.5 Flash: benchmark scores, pricing & comparison.
Side-by-side Sakana Fugu Ultra vs Gemini 3.5 Flash comparison across SWE-bench, GPQA, HLE, Terminal-Bench, coding agent scores, token pricing, context window, and AskClash RWT. Green marks the winner on each benchmark.
Rank #1 vs #15AskClash overall scores 96.5 vs 58.2.
Pricing —/— vs $1.50/$9.00Input and output token prices per 1M tokens when published.
Proprietary vs ProprietarySakana AI vs Google.
Sakana Fugu Ultra vs Gemini 3.5 Flash benchmark comparison
Green cells highlight the winning model for each metric. Scores are cached from the AskClash LLM leaderboard snapshot.
| Metric | Sakana Fugu Ultra | Gemini 3.5 Flash |
|---|---|---|
| Overall Score | 96.5 | 58.2 |
| Leaderboard Rank | #1 | #15 |
| RWT | — | 7.0 |
| HLE | — | 40.2 |
| GPQA | 95.5 | 92.2 |
| IFEval | — | 76.3 |
| SWE-Pro | — | 55.1 |
| Terminal-Bench | 82.1 | 76.2 |
| LiveCodeBench | 93.2 | — |
| OSWorld | — | 78.4 |
| MCP Atlas | — | 83.6 |
| Finance Agent | — | 57.9 |
| CharXiv | — | 84.2 |
| MMMU-Pro | — | 83.6 |
| ARC-AGI 2 | — | 72.1 |
| Tau2 | — | 95.3 |
| MRCR | — | 77.3 |
| Input Price (per 1M tokens) | — | $1.50 |
| Output Price (per 1M tokens) | — | $9.00 |
| Context Window | 1M | 1M |
| Benchmark Cells | 3 | 14 |
More Sakana Fugu Ultra and Gemini 3.5 Flash comparisons
Explore how Sakana Fugu Ultra and Gemini 3.5 Flash stack up against other top-ranked LLMs.
Sakana Fugu Ultra vs Claude Mythos Fable 5Claude Mythos Fable 5 vs Gemini 3.5 FlashSakana Fugu Ultra vs Claude Opus 4.8Claude Opus 4.8 vs Gemini 3.5 FlashSakana Fugu Ultra vs GLM-5.2GLM-5.2 vs Gemini 3.5 FlashSakana Fugu Ultra vs GPT-5.5GPT-5.5 vs Gemini 3.5 FlashSakana Fugu Ultra vs Claude Sonnet 5Claude Sonnet 5 vs Gemini 3.5 Flash
How to read this comparison
Token pricing
Lower is better for input and output prices. Green marks the cheaper model per 1M tokens.
Coverage matters
Models with fewer disclosed benchmark cells may have inflated percentile scores. Check the benchmark cell count for context.
This comparison page is generated from the AskClash LLM leaderboard cache. Open the live leaderboard for real-time scores and interactive filtering.