ArabCulture-Dialogue - Dialect to MSA Translation (BLEU): leaderboard

ArabCulture-Dialogue (MBZUAI, 2026) asks models to translate culturally grounded dialogues from an Arab country dialect into Modern Standard Arabic, zero-shot with the target country and region named in the prompt. The score is BLEU-4 against the human reference translation on a 0-100 scale; higher is better.

Metric: BLEU-4 (0-100, the paper's 0-1 score times 100) against the human reference, zero-shot Dialect to MSA translation of the dialogues of ArabCulture-Dialogue (13 Arab countries, parallel Modern Standard Arabic and country-dialect versions of 3,471 culturally grounded multi-turn dialogues built from ArabCulture), with the target country and region named in the prompt; higher is better. Source: arxiv.org. Saturation forecast: Around March 2028. 11 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-543.4#178
2Gemini 2.5 Pro43#276
3ALLaM-7B-Instruct-preview40.5#2001
4Fanar-1-9B36.7#1913
5Gemma 2 9B (IT)31.9#2369
6Hala-9B31.5#1162
7c4ai-command-r7B-arabic-02-202531#2135
8Qwen 3 8B26.5#1873
9Llama 3.1 8B Instruct24.2#4600
10SILMA-9B-Instruct-v1.021.8#3847
11jais-adapted-7B-chat17.8#5831

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=arabculture-dialogue-dialect-to-msa-translation-bleu · How It Works · Data refreshed daily, snapshot 2026-11-21.