Gemini 2.5 Flash vs GPT-4.1

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

Shared benchmarks
2
Gemini 2.5 Flash only
3
GPT-4.1 only
2

Shared benchmarks, side by side

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

  • Gemini 2.5 Flash
  • GPT-4.1

  • FAITH overall accuracy %, higher is better

    Gemini 2.5 Flash:88.7
    GPT-4.1:89.2 (better)

  • FinTrust safety harmfulness score (1-10, GPT-4.1 mini judge), lower is better

    Gemini 2.5 Flash:6.2 (better)
    GPT-4.1:7.4
Benchmark
Gemini 2.5 Flash
GPT-4.1

7 benchmarks, 2 scored for both. Bold marks the better score where both have one.

FAITHoverall accuracy %, higher is better
88.7
Paper Table 3 (Reported by the authors, 24 Oct 2025)
89.2
Paper Table 3 (Reported by the authors, 24 Oct 2025)
FinTrustsafety harmfulness score (1-10, GPT-4.1 mini judge), lower is better
6.2
Paper Table 7 (App. A.2) (Reported by the authors, 17 Oct 2025)
7.4
Paper Table 7 (App. A.2) (Reported by the authors, 17 Oct 2025)
AI-Tradercumulative return %, higher is better
-0.060
Paper Table 1 (Reported by the authors, 1 Dec 2025)
Not scored
CAIA (Crypto AI Agent Benchmark)accuracy % (majority vote, 5 runs), higher is better
Not scored
46.6
Paper Tables 2-3 (Reported by the authors, 17 Jan 2026)
Advanced Financial Reasoning at Scale (CFA Level III evaluation)overall % (mean of MCQ and essay), higher is better
77.3
Paper Table 1 (Reported by the authors, 22 Sept 2025)
Not scored
K-FinHallubinary detection F1, higher is better
0.860
Paper Table 3 (Reported by the authors, 28 May 2026)
Not scored
LiveTradeBenchcumulative return %, higher is better
Not scored
6.3
Paper Table 2 (Reported by the authors, 5 Nov 2025)

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