New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
FinanceReasoning benchmark
AI model leaderboard for the FinanceReasoning benchmark. Compare how large language models score on FinanceReasoning, see the full ranking, and understand what this AI benchmark measures. GPT-5 currently leads with 89.33. Quantitative finance reasoning benchmark — multi-step problems involving valuation, risk, and accounting.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | GPT-5 | OpenAI | 89.33 | thinking-high-cited-ring-1t | official ↗ |
| #2 | DeepSeek-V3.1-Terminus | DeepSeek | 87.76 | thinking-cited-ring-1t | official ↗ |
| #3 | Qwen3-235B-A22B-Thinking (Jul 2025) | Qwen | 87.65 | thinking-cited-ring-1t | official ↗ |
| #4 | Ling-1T | Ant Group | 87.45 | official ↗ | |
| #5 | Ring-1T | Ant Group | 87.42 | thinking | official ↗ |
| #6 | gemini-2.5-pro | Google DeepMind | 87.33 | cited-ring-1t | official ↗ |
| #7 | gemini-2.5-pro | Google DeepMind | 86.65 | lowthink-cited-ling-1t | official ↗ |
| #8 | DeepSeek-V3.1-Terminus | DeepSeek | 86.44 | non-thinking-cited-ling-1t | official ↗ |
| #9 | GPT-5 | OpenAI | 86.28 | main-cited-ling-1t | official ↗ |
| #10 | Kimi K2 | Moonshot | 84.83 | instruct-0905-cited-ling-1t | official ↗ |
Frequently asked questions about FinanceReasoning
What is the FinanceReasoning benchmark?
Quantitative finance reasoning benchmark — multi-step problems involving valuation, risk, and accounting.
How is the FinanceReasoning benchmark scored?
FinanceReasoning is scored using the accuracy metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on FinanceReasoning?
As of the latest reported scores on GenAIList, GPT-5 achieves the highest result on FinanceReasoning with a score of 89.33.
Is a higher FinanceReasoning score better?
Yes. On FinanceReasoning a higher score indicates better performance, so models near the top of the leaderboard are the strongest.