Comparing 2 models
Claude Opus 5.5 vs Gemini 4 Argon
Benchmarks through
Verdict
Claude Opus 5.5 places higher on 6 of the 11 tasks where both are ranked. Gemini 4 Argon costs 2.0x less per input token. Gemini 4 Argon accepts 1.0x more context.
- Choose Claude Opus 5.5 for Math & Reasoning (#1), Competition Math (#1), Coding (#2).
- Choose Gemini 4 Argon for Logic & Problem Solving (#2), a General Assistant (#2), Massive Context (#2).
Built from the published task rankings below; ranks come from each task's own weighted benchmarks, so a win on one task says nothing about another.
Measured head to head
Benchmarks both models were run on
| Evaluation | Claude Opus 5.5 | Gemini 4 Argon | Source |
|---|---|---|---|
| Design Arena Coding | 1385 Elo | 1326 Elo | source · 2026-11-15 |
| Design Arena UI | 1375 Elo | 1331 Elo | source · 2026-11-15 |
| Design Arena Web-app Agent | 1321 Elo | 1283 Elo | source · 2026-11-15 |
| Design Arena Website | 1326 Elo | 1316 Elo | source · 2026-11-15 |
| FACTS Grounding | 62.4% | 73.4% | source · 2026-10-24 |
| FrontierMath Tiers 1-3 (Epoch AI) | 91.2% ± 1.7 | 86% ± 2.1 | source · 2026-09-22 |
| GPQA Diamond (Epoch AI) | 90.6% ± 1.9 | 96.4% ± 1.2 | source · 2026-09-22 |
| LMArena Agent | 19.5 | 10.7 | source · 2026-11-11 |
| LMArena Agent Bash Recovery | 13.6 | -21.4 | source · 2026-10-08 |
| LMArena Agent Steerability | 18.6 | 13.2 | source · 2026-11-11 |
| LMArena Agent Task Outcome | 20.7 | 13.2 | source · 2026-11-11 |
| LMArena Agent Tool Hallucination | 0.7 | 0.5 | source · 2026-11-11 |
| LMArena Creative Writing | 1532 Elo | 1533 Elo | source · 2026-11-11 |
| LMArena Instruction Following | 1535 Elo | 1536 Elo | source · 2026-11-11 |
| LMArena Long Query | 1528 Elo | 1547 Elo | source · 2026-11-11 |
| LMArena Math | 1520 Elo | 1544 Elo | source · 2026-11-11 |
| LMArena Text | 1507 Elo | 1522 Elo | source · 2026-11-11 |
| LMArena Text Factuality | 1505 Elo | 1497 Elo | source · 2026-11-11 |
| LMArena WebDev | 1815 Elo | 1681 Elo | source · 2026-11-11 |
| OTIS Mock AIME (Epoch AI) | 100% ± 0 | 100% ± 0 | source · 2026-09-22 |
| SimpleQA Verified (Epoch AI) | 72.2% ± 1.4 | 76% ± 1.4 | source · 2026-09-22 |
| Terminal-Bench (4.1) | 65.8% | 55.5% | source · 2026-11-15 |
Values are as published by each evaluator, in its own units; rows from different evaluators are not on a common scale.
task evidence
How they place when the question changes
Lower ranks are better. “Best” marks the strongest option in this shortlist for that task.