FinReasoning (Semantic Consistency) - Fact: leaderboard
Metric: Composite score (0-100): the mean of sentence-level F1 for locating the injected errors, an error-explanation score and a corrected-content score, each of the last two the mean of BERTScore, SimCSE similarity and a judge score fused from DeepSeek-V3 and Qwen3-235B against the expert-verified annotation, on the 600 factual-error items (relation, entity and context errors) of FinReasoning's Semantic Consistency track (long financial passages with one to three injected errors to locate, explain and correct); Chinese-language tasks built from A-share research reports, news and market data (2023 to 2025); zero-shot at temperature 0.1; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 19 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Seed 1.8 | 74.4 | #136 |
| 2 | GLM-4.6 | 73.4 | #246 |
| 3 | GPT-5 | 71.4 | #91 |
| 4 | Gemini 3 Pro | 71.1 | #77 |
| 5 | Qwen 3 Max | 70 | #201 |
| 6 | Claude Sonnet 4.5 | 69 | #138 |
| 7 | DeepSeek R1 | 66.9 | #245 |
| 8 | Kimi K2 | 66 | #236 |
| 9 | DeepSeek V3 | 63.1 | #312 |
| 10 | Qwen 3 235B A22B | 57.7 | #304 |
| 11 | Intern-S1 | 57.1 | #278 |
| 12 | Qwen 3 32B | 54.9 | #424 |
| 13 | GPT-4o | 53.8 | #333 |
| 14 | Qwen 3 8B | 46.1 | #667 |
| 15 | Llama 3.1 70B | 28.8 | #578 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=finreasoning-semantic-consistency-fact · How It Works · Data refreshed daily, snapshot 2026-10-11.