CommunityFact (Closed-Input) - Finance (Japanese): leaderboard
Metric: Macro-F1 (%) over True and False claims on the temporally held-out Finance Japanese test slice (80 True and 47 False standalone claims built from helpful X Community Notes, labels grounded in the note), closed-input (the model sees only the claim and its timestamp, no web search), zero-shot verdict prompt in the claim language; open models greedy, closed models at default settings with reasoning effort medium in their thinking runs; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash (Non-reasoning) | 81.6 |
| 2 | aya-expanse-32B | 79.13 |
| 3 | Grok 4.3 (Medium) | 78.48 |
| 4 | Gemini 2.5 Flash (Medium) | 78.48 |
| 5 | aya-expanse-8B | 74.97 |
| 6 | Qwen 3 32B (Non-reasoning) | 73.03 |
| 7 | EuroLLM-22B-Instruct-2512 | 66.88 |
| 8 | Qwen 3 32B | 66.11 |
| 9 | Qwen 3 14B (Non-reasoning) | 64.53 |
| 10 | Qwen 3 14B | 62.84 |
| 11 | GPT-5 Nano (Medium) | 61.8 |
| 12 | Grok 4.3 (Non-reasoning) | 52.11 |
| 13 | GPT-5 Nano (Minimal) | 40.35 |
| 14 | Ministral-8B-Instruct-2410 | 34.44 |
Interactive version: theaggregate.ai/benchmark?slug=communityfact-closed-input-finance-japanese · How It Works · Data refreshed daily, snapshot 2026-10-07.