Hebrew LLM - Trivia (0-shot): leaderboard
The Trivia task of the Hebrew LLM Leaderboard — zero-shot Hebrew trivia question answering; accuracy (%).
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around January 2027. 34 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 86.71 | #66 |
| 2 | Gemini 3 Flash (Preview) | 86.38 | #82 |
| 3 | Gemini 2.5 Pro | 82.06 | #152 |
| 4 | GPT-5.1 | 81.73 | #135 |
| 5 | GPT-4o | 77.08 | #343 |
| 6 | GLM-5.1 | 76.74 | #97 |
| 7 | GLM-5.3 Flash | 76.67 | #65 |
| 8 | Gemini 2.5 Flash | 76.41 | #245 |
| 9 | GPT-OSS-120B | 74.82 | #341 |
| 10 | Qwen 3.5 397B A17B | 72.09 | #146 |
| 11 | Qwen 3.5 122B A10B | 70.1 | #173 |
| 12 | GPT-5 Mini | 69.44 | #186 |
| 13 | Claude Sonnet 4.5 | 69 | #142 |
| 14 | jamba-large-1.7 | 66.11 | #525 |
| 15 | DeepSeek V3.1 | 64.45 | #264 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=hebrew-llm-trivia-0-shot · How It Works · Data refreshed daily, snapshot 2026-10-20.