CulturalRiddles - Exact Match: leaderboard
Deterministic rule-based exact-match accuracy (%) on Cohere Labs Community's CulturalRiddles benchmark across 57 human-evaluated communities prior to human and LLM-jury review.
Metric: Rule-based exact-match accuracy (%, 57 communities). Source: huggingface.co. Saturation forecast: Around February 2027. 25 models tracked. Updated 20 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 48.5 | #101 | |
| 2 | GPT-5.5 | Non-reasoning | 42.8 | #54 |
| 3 | DeepSeek V4 Pro | High | 36.6 | #198 |
| 4 | Claude Opus 4.7 | High | 34.2 | #85 |
| 5 | Gemini 2.5 Flash | 25.3 | #469 | |
| 6 | DeepSeek V4 Flash | 21.5 | #267 | |
| 7 | Mistral Large 3 | 20.6 | #835 | |
| 8 | Gemma 4 31B (IT) | 20.4 | #354 | |
| 9 | Qwen-SEA-LION-v4.5-27B-IT | 17.7 | #603 | |
| 10 | Gemini 2.5 Flash Lite | 17.4 | #974 | |
| 11 | Qwen 3.5 27B | 13.8 | #378 | |
| 12 | Apertus-v1.5-70B | 9.5 | #1783 | |
| 13 | EuroLLM-22B-Instruct-2512 | 7.6 | #2012 | |
| 14 | Apertus-70B-Instruct-2509 | 7.1 | #2627 | |
| 15 | Nova Lite | 6.4 | #2322 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=culturalriddles-exact-match · How It Works · Data refreshed daily, snapshot 2026-11-20.