UGI - Natural Intelligence: leaderboard
The Natural Intelligence column of the Uncensored General Intelligence leaderboard, combining textbook knowledge such as history, statistics, math and logic, pop culture detail, and world model tests such as estimating recipe quantities and object weights, placing a described location, and predicting a viewer's show ratings. All test questions are kept private. Scored 0 to 100, higher is better.
Metric: NatInt Score. Source: huggingface.co. Saturation forecast: Around January 2027. 1329 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) (High) | 74.94 | #58 (Gemini 3.1 Pro (Preview)) |
| 2 | Gemini 3 Pro (Preview) (High) | 74.29 | #66 (Gemini 3 Pro (Preview)) |
| 3 | Gemini 3 Flash (Preview) (Medium) | 72.42 | #82 (Gemini 3 Flash (Preview)) |
| 4 | Gemini 3.5 Flash (Minimal) | 72.31 | #60 (Gemini 3.5 Flash) |
| 5 | Claude Opus 4.6 (Adaptive Reasoning, Max Effort) | 70.78 | #63 (Claude Opus 4.6) |
| 6 | Gemini 3 Flash (Preview) (Minimal) | 70.41 | #82 (Gemini 3 Flash (Preview)) |
| 7 | GPT-5.6 Sol (Non-reasoning) | 70.1 | #19 (GPT-5.6 Sol) |
| 8 | Gemini 3.5 Flash (High) | 68.98 | #60 (Gemini 3.5 Flash) |
| 9 | Gemini 3 Flash (Preview) (High) | 68.85 | #82 (Gemini 3 Flash (Preview)) |
| 10 | DeepSeek V4 Pro (Reasoning) | 67.03 | #102 (DeepSeek V4 Pro) |
| 11 | O3 (2025-04-16) (High) | 66.19 | #121 (O3 (2025-04-16)) |
| 12 | O3 (2025-04-16) (Medium) | 65.92 | #121 (O3 (2025-04-16)) |
| 13 | O3 (Low) | 65.53 | #124 (O3) |
| 14 | Kimi K2.6 (Thinking) | 65.25 | #104 (Kimi K2.6) |
| 15 | Claude Opus 4.5 (20251101) (Thinking) | 65.22 | #74 (Claude Opus 4.5 (20251101)) |
No result here: #8 Gemini 4 Argon, #10 Muse Wonder Full.
Interactive version: theaggregate.ai/benchmark?slug=ugi-natural-intelligence · How It Works · Data refreshed daily, snapshot 2026-10-20.