# Claude Opus 5 vs GPT-5.5

Every finance benchmark where both have a score (Anthropic and OpenAI). Data refreshed 2 Nov 2026. Scores from different benchmarks use different scales and are not added up.

| Benchmark | Metric | Claude Opus 5 | Claude Opus 5 source | GPT-5.5 | GPT-5.5 source | Ahead |
| --- | --- | --- | --- | --- | --- | --- |
| τ³-Banking | pass^1 % (higher is better) | 48.7 | τ-bench leaderboard (Sierra), 4 Aug 2026, https://taubench.com/leaderboard/?benchmark=knowledge | 44.6 | τ-bench leaderboard (Sierra), 5 May 2026, https://taubench.com/leaderboard/?benchmark=knowledge | Claude Opus 5 |
| CorpFin v2 | accuracy % (higher is better) | 73.2 | Vals.ai leaderboard (archived), 12 Aug 2026, https://www.vals.ai/benchmarks/corp_fin_v2 | 68.4 | Vals.ai leaderboard (archived), 12 Aug 2026, https://www.vals.ai/benchmarks/corp_fin_v2 | Claude Opus 5 |

Source: https://benchmarks.finance/compare/claude-opus-5-vs-gpt-5-5/

Created and maintained by Moneda.