Gapminder AI Worldview — leaderboard
Metric: Correct Rate (%). Source: www.gapminder.org. 37 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 (Medium) | 93.9 |
| 2 | O1 (Medium) | 90.9 |
| 3 | GPT-4.1 | 89.1 |
| 4 | Claude 3.5 Sonnet | 88.6 |
| 5 | Claude 3.7 Sonnet (Thinking) | 87.9 |
| 6 | GPT-4o | 85 |
| 7 | Claude Sonnet 4 (Thinking) | 84.6 |
| 8 | Grok 3 | 82.3 |
| 9 | GPT-4 Turbo | 81.5 |
| 10 | Gemini 1.5 Pro (002) | 80.9 |
| 11 | Sonar | 78.4 |
| 12 | Llama 4 Maverick Instruct | 78.3 |
| 13 | DeepSeek V3 | 78.1 |
| 14 | Llama 3.3 70B Instruct | 75 |
| 15 | O3 Mini (High) | 73.7 |
Interactive version: theaggregate.ai/benchmark?slug=gapminder-ai-worldview · How the rankings work · Data refreshed daily, snapshot 2026-07-22.