benchmarks.site

Public leaderboard · Polish language

Polish Linguistic and Cultural Competency (PLCC)

600 hand-written questions about Polish history, geography, culture and traditions, art and entertainment, grammar and vocabulary. It shows how well an AI model knows Poland and the Polish language, not just how fluent it sounds.

Leaderboard

Top 3 right now

  1. 1stGemini-4-Carbon98.33
  2. 2ndGemini-4-Argon97.33
  3. 3rdClaude-Fable-5.297.33

What it measures

600 hand-written questions in six categories (history, geography, culture and tradition, art and entertainment, grammar, vocabulary), scored by accuracy.

How to read the numbers

Share of the 600 questions answered fully correctly (%), averaged over the six categories — higher is better.

Results

25 rows · higher is better (Overall)CSV
1Gemini-4-CarbonGoogle98.3397100100979898
2Gemini-4-ArgonGoogle97.33959899969898
3Claude-Fable-5.2Anthropic97.33969898979996
4Gemini-3.1-Pro-PreviewGoogle9795100100939896
5Gemini-3.8-FlashGoogle979599100959895
6GPT-6.1-SolOpenAI96.83939898989995
7Gemini-4-HydrogenGoogle96.839397100979797
8Gemini-3.7-FlashGoogle96.679598100959795
9Gemini-3.5-FlashGoogle96.3394100100979691
10GPT-6-AstraOpenAI96.33949898959994
11Gemini-3.0-Pro-PreviewGoogle95.839599100919595
12GPT-5.5 (high reasoning)OpenAI95.83939898979594
13Gemini-3.6-FlashGoogle95.5929899949694
14Claude-Fable-5.1Anthropic95.17919599959794
15Claude-Fable-5Anthropic95939698949792
16GPT-5.6-Sol (medium reasoning)OpenAI94.67939699919693
17GPT-5.5 (low reasoning)OpenAI94.5939699919593
18GPT-6.1-TerraOpenAI94909497939892
19GPT-6-Sol (medium reasoning)OpenAI93.83949398939788
20Muse-WonderMeta93.83929298929792
21Claude-Opus-5.5Anthropic92.5869394949791
22Gemini-2.5-Pro-Preview-06-05Google92.17919698869290
23GPT-5.4-2026-03-05 (high reasoning)OpenAI92.17919396909291
24Claude-Opus-5Anthropic92.17869497929688
25Muse-Spark-1.3Meta91.83899497899488

Note: Per-category accuracies copied from the Space's data.json (253 models). The leaderboard file does not include an overall column; 'Overall' here is the plain mean of the six category scores (each category has 100 questions, so this equals the overall percentage of correct answers), and rows are ranked by it. as_of is the Space's last commit date.

Source: Sławomir Dadas et al. (OPI PIB) · huggingface.co · results as of 2026-11-16 · retrieved 2026-11-17 · license: CC-BY-4.0

Markdown · More Polish language