HalluScore - History: leaderboard

HalluScore (2026) is an Arabic question-answering benchmark of 827 hallucination-prone Modern Standard Arabic questions, with human annotation of model answers. This board is the factual hallucination rate, the percent of answers that state an unsupported fact with confidence on the 240 history questions, zero-shot at temperature 0 (expressed uncertainty or refusal does not count); lower is better.

Metric: Factual hallucination rate (%) on the 240 historical HalluScore questions, zero-shot short-answer Arabic question answering at temperature 0 with no retrieval, responses annotated by humans; an answer that states an unsupported fact with confidence counts, an expressed uncertainty or refusal does not; lower is better. Source: arxiv.org. Saturation forecast: Around February 2027. 17 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1GPT-529.17#185
2Claude Opus 437.5#332
3GPT-4o42.08#754
4Claude Sonnet 4.542.5#267
5O4 Mini52.92#372
6DeepSeek V355.42#736
7Llama 4 Maverick Instruct FP856.25#1034
8Grok 4 FastReasoning57.5#446
9DeepSeek R161.67#484
10Qwen3-235B-A22B-Instruct-2507-tput67.08#557
11Qwen 3 Next 80B A3B Instruct67.92#781
12ALLaM-7B-Instruct-preview67.92#1992
13Grok 470.83#321
14HalluScore Jais-6.7B (checkpoint unspecified)77.92
15O383.33#242

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=halluscore-history · How It Works · Data refreshed daily, snapshot 2026-11-22.