MORPHOGEN - Arabic Corpus Gender Accuracy: leaderboard

Gender-aware morphological generation in Arabic on MORPHOGEN (2026): the model rewrites a first-person sentence for a speaker of the opposite grammatical gender while keeping its meaning, scored against human-corrected references. Corpus-level share of reference gendered terms generated correctly, in percent; higher is better.

Metric: Corpus-level gender accuracy (%): correctly generated gendered terms over all reference gendered terms of the test set, in the GENFORM task of MORPHOGEN: rewrite a first-person Arabic sentence for a speaker of the opposite grammatical gender while keeping its meaning (both directions, human-corrected references), zero-shot with the model chat template, temperature 0.1; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 15 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-4o Mini80.27#1667
2Gemma 3 27B74.74#1663
3Phi-466.15#1931
4Gemma 3 12B65.52#2059
5Llama 3.3 70B64.37#1514
6Qwen 3 14B56.08#1269
7Qwen 3 32B53#972
8Qwen 3 8B51.01#1873
9Gemma 3 4B48.93#5339
10Llama 3.1 8B45.51#5196
11Gemma 2 9B45.26#3201
12Qwen 3 4B35.97#2348
13Llama 3.2 3B17.75#7877
14Gemma 2 2B14.1#8415
15Qwen 2.5 0.5B4.59#9144

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

Interactive version: theaggregate.ai/benchmark?slug=morphogen-arabic-corpus-gender-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-21.