HalluScore - Reasoning: 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 220 questions that require reasoning, zero-shot at temperature 0 (expressed uncertainty or refusal does not count); lower is better.

Metric: Factual hallucination rate (%) on the 220 HalluScore questions that require reasoning, 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-520.91#185
2Claude Opus 426.82#332
3Claude Sonnet 4.530.91#267
4O4 Mini35.91#372
5Grok 4 FastReasoning42.73#446
6GPT-4o43.18#754
7DeepSeek V346.82#736
8DeepSeek R150.91#484
9Llama 4 Maverick Instruct FP851.36#1034
10Qwen 3 Next 80B A3B Instruct60.91#781
11Qwen3-235B-A22B-Instruct-2507-tput61.36#557
12Grok 462.73#321
13ALLaM-7B-Instruct-preview68.64#1992
14O377.73#242
15Fanar-1-9B80.91#1925

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

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