Claude Opus 4.1 vs DeepSeek-V3.1 in finance

Every finance benchmark where both have a score (Anthropic and DeepSeek). Data refreshed 2 Nov 2026.

Shared benchmarks
2
Claude Opus 4.1 only
1
DeepSeek-V3.1 only
2

Shared benchmarks, side by side

Each benchmark is drawn on its own scale. Bars are not comparable across rows.

  • Claude Opus 4.1
  • DeepSeek-V3.1
Benchmark
Claude Opus 4.1
DeepSeek-V3.1

6 benchmarks, 3 scored for both. Bold marks the better score where both have one.

CAIA (Crypto AI Agent Benchmark)accuracy % (majority vote, 5 runs), higher is better
56.3
Paper Tables 2-3 (Reported by the authors, 17 Jan 2026)
49.2
Paper Tables 2-3 (Reported by the authors, 17 Jan 2026)
Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation)Level I accuracy %, higher is better
94.6
Paper Table 5 (Reported by the authors, 9 Dec 2025)
90.9
Paper Table 5 (Reported by the authors, 9 Dec 2025)
Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation)Level III constructed-response score %, higher is better
73.4
Paper Table 5 (Reported by the authors, 9 Dec 2025)
70.8
Paper Table 5 (Reported by the authors, 9 Dec 2025)
AI-Tradercumulative return %, higher is better
Not scored
8.4
Paper Table 1 (Reported by the authors, 1 Dec 2025)
LiveTradeBenchcumulative return %, higher is better
3.7
Paper Table 2 (Reported by the authors, 5 Nov 2025)
Not scored
StockBenchfinal return %, higher is better
Not scored
1.1
Paper Table 2 (Reported by the authors, 2 Mar 2026)

Scores from different benchmarks use different scales and are not added up. Settings can differ between sources; see each leaderboard for details.