Adam Mickiewicz University's
Center for Artificial Intelligence

LLMzSzŁ: a comprehensive LLM benchmark for Polish

Overall scores

LLMzSzŁ (LLM-y za Szkolną Ławą, LLMs Behind the School Desk) is a comprehensive LLM benchmark for Polish, based on a collection of Polish national exams, including both academic and professional tests extracted from the archives of the Polish Central Examination Board.

Select columns to display

Name Lang Score Parameters (B)

Name	Lang	Score	Parameters (B)
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	Multilingual	69.06	46.7

Name	Lang	Score	Parameters (B)
Qwen/Qwen2.5-72B-Instruct	English	69.06	72
Qwen/Qwen2.5-72B	English	68.5	72
mistralai/Mistral-Large-Instruct-2407	Multilingual	67.17	123
meta-llama/Llama-3.3-70B-Instruct	Multilingual	67.13	70
Qwen/Qwen3-32B	English	66.84	32
mistralai/Mistral-Large-Instruct-2411	Multilingual	66.61	123
meta-llama/Meta-Llama-3.1-70B-Instruct	Multilingual	66.59	70
CYFRAGOVPL/Llama-PLLuM-70B-base	Polish	64.56	70
CYFRAGOVPL/Llama-PLLuM-70B-chat	Polish	64.42	70
meta-llama/Meta-Llama-3-70B-Instruct	Multilingual	64.04	70
Qwen/Qwen3-30B-A3B	English	63.52	30
meta-llama/Llama-3.1-70B	English	63.12	70
meta-llama/Meta-Llama-3-70B	Multilingual	62.22	70
Qwen/Qwen2.5-32B	English	61.04	32
Qwen/Qwen3-14B	English	61.02	14
CYFRAGOVPL/PLLuM-8x7B-nc-chat	Polish	60.52	46.7
speakleash/Bielik-11B-v2.6-Instruct	Polish	59.23	11
speakleash/Bielik-11B-v2.5-Instruct	Polish	58.93	11
mistralai/Mixtral-8x22B-Instruct-v0.1	Multilingual	58.17	141
CYFRAGOVPL/PLLuM-8x7B-nc-base	Polish	57.82	46.7
speakleash/Bielik-11B-v2.1-Instruct	Polish	57.52	11
speakleash/Bielik-11B-v2.3-Instruct	Polish	57.4	11
speakleash/Bielik-11B-v2.2-Instruct	Polish	57.36	11
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	English	55.8	32
speakleash/Bielik-11B-v2.0-Instruct	Polish	55.61	11
Qwen/Qwen2.5-14B	English	55.31	14
Qwen/Qwen3-8B	English	55.25	8
speakleash/Bielik-11B-v2	Polish	55.14	11
speakleash/Bielik-4.5B-v3.0-Instruct	Polish	54.76	4.5
CYFRAGOVPL/PLLuM-12b-nc-chat-250715	Polish	54.24	12
CYFRAGOVPL/PLLuM-12B-nc-chat	Polish	53.4	12
CYFRAGOVPL/PLLuM-8x7B-chat	Polish	52.8	46.7
CYFRAGOVPL/PLLuM-8x7B-base	Polish	52.24	46.7
microsoft/phi-4	English	51.76	14.7
utter-project/EuroLLM-9B-Instruct	Multilingual	51.35	9
CYFRAGOVPL/PLLuM-12B-nc-base	Polish	50.55	12
Qwen/Qwen3-4B	English	50.2	4
CYFRAGOVPL/PLLuM-12B-chat	Polish	49.69	12
mistralai/Mixtral-8x7B-Instruct-v0.1	Multilingual	49.46	46.7
mistralai/Mixtral-8x7B-v0.1	Multilingual	47.71	46.7
CYFRAGOVPL/Llama-PLLuM-8B-chat	Polish	47.68	8
meta-llama/Meta-Llama-3.1-8B-Instruct	Multilingual	47.41	8
google/gemma-7b	English	46.84	7
speakleash/Bielik-1.5B-v3.0-Instruct	Polish	46.81	1.5
deepseek-ai/DeepSeek-R1-Distill-Qwen-14B	English	46.78	14
CYFRAGOVPL/PLLuM-12B-base	Polish	46.47	12
meta-llama/Llama-2-70b-hf	English	46.38	70
CYFRAGOVPL/Llama-PLLuM-8B-base	Polish	46.32	8
Qwen/Qwen2-7B	English	45.59	7
mistralai/Mistral-Nemo-Instruct-2407	Multilingual	45.54	12.2
meta-llama/Meta-Llama-3-8B-Instruct	Multilingual	44.83	8
meta-llama/Meta-Llama-3.1-8B	Multilingual	44.21	8
allenai/OLMo-2-0325-32B-Instruct	English	44.18	32
mistralai/Ministral-8B-Instruct-2410	Multilingual	42.27	8
mistralai/Mistral-Nemo-Base-2407	Multilingual	42.16	12.2
allenai/OLMo-2-0325-32B	English	41.83	32
mistralai/Mistral-7B-Instruct-v0.3	Multilingual	41.72	7
01-ai/Yi-1.5-34B-Chat	Multilingual	41.47	34
01-ai/Yi-34B-Chat	Multilingual	41.42	34
meta-llama/Meta-Llama-3-8B	Multilingual	41.38	8
01-ai/Yi-1.5-34B	Multilingual	41.14	34
speakleash/Bielik-7B-Instruct-v0.1	Polish	40.77	7
mistralai/Mistral-7B-Instruct-v0.2	Multilingual	40.75	7
01-ai/Yi-34B-Chat-4bits	Multilingual	40.28	34
Voicelab/trurl-2-13b-8bit	Polish	40.23	13
Voicelab/trurl-2-13b	Polish	40.22	13
Qwen/Qwen3-1.7B	English	39.89	1.7
meta-llama/Llama-3.2-3B-Instruct	Multilingual	39.86	3
speakleash/Bielik-7B-v0.1	Polish	39.15	7
lucyknada/microsoft_WizardLM-2-7B	Multilingual	38.23	7
mistralai/Mistral-7B-v0.1	Multilingual	37.75	7
01-ai/Yi-1.5-9B-Chat	Multilingual	37.59	9
01-ai/Yi-34B-200K	Multilingual	37.56	34
mistralai/Mistral-7B-v0.3	Multilingual	37.08	7
01-ai/Yi-1.5-9B	Multilingual	37.06	9
mistralai/Mistral-7B-Instruct-v0.1	Multilingual	35.98	7
Voicelab/trurl-2-13b-academic	Polish	34.89	13
OPI-PG/Qra-13b	Polish	34.85	13
Qwen/Qwen2-1.5B	English	34.19	5
meta-llama/Llama-2-13b-hf	English	33.85	13
microsoft/Phi-3-mini-4k-instruct	English	33.44	4
allenai/OLMo-2-1124-13B-Instruct	English	33.23	13
allenai/OLMo-2-1124-13B	English	32.61	13
Voicelab/trurl-2-7b	Polish	32.3	7
Voicelab/trurl-2-7b-8bit	Polish	31.86	7
meta-llama/Llama-3.2-1B-Instruct	Multilingual	31.6	1
allenai/OLMo-2-1124-7B-Instruct	English	31.5	7
allenai/OLMo-2-1124-7B	English	31.09	7
OPI-PG/Qra-7b	Polish	29.07	7
meta-llama/Llama-2-7b-hf	English	28.04	7
microsoft/phi-1	English	25.73	1
utter-project/EuroLLM-1.7B-Instruct	Multilingual	25.61	1.7
microsoft/phi-2	English	25.6	3
OPI-PG/Qra-1b	Polish	25.47	1
sdadas/polish-gpt2-large	Polish	24.89	0.9
sdadas/polish-gpt2-medium	Polish	24.4	0.5
microsoft/phi-1_5	English	24.25	1
sdadas/polish-gpt2-small	Polish	24.19	0.2
sdadas/polish-gpt2-xl	Polish	23.98	2

Middle School Exam

The Middle School Exam (egzamin gimnazjalny) was formerly a key component of Poland’s secondary education system. This exam was taken by students completing their lower secondary education.

Subjects Covered

This exam assessed students in core subjects like:

Polish Language
Mathematics
Science
Foreign Language

preparing them for higher levels of education.

Recent Changes

Due to recent educational reforms in Poland:

Middle Schools Phased Out: Middle schools have been removed from the education system.
Exam Discontinued: With this change, the Middle School Exam has also been discontinued.

Model	2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
speakleash/Bielik-11B-v2.3-Instruct_newer	43.75	71.43	46.67	33.33	33.33	44.44	12.5	85.71	22.22	81.25	33.33	37.5	12.5	11.11	40	40	57.14	33.33

Model	2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
01-ai/Yi-1.5-34B-Chat	37.50	71.43	46.67	60.00	33.33	44.44	12.50	85.71	22.22	50.00	33.33	37.50	25.00	11.11	40.00	40.00	57.14	33.33
01-ai/Yi-1.5-34B	43.75	71.43	46.67	60.00	50.00	66.67	25.00	71.43	44.44	81.25	33.33	50.00	25.00	66.67	40.00	30.00	57.14	44.44
01-ai/Yi-1.5-9B-Chat	31.25	71.43	26.67	40.00	50.00	22.22	50.00	85.71	44.44	37.50	16.67	25.00	50.00	55.56	30.00	40.00	28.57	44.44
01-ai/Yi-1.5-9B	50.00	71.43	26.67	33.33	66.67	33.33	37.50	57.14	55.56	31.25	16.67	12.50	12.50	44.44	20.00	40.00	42.86	33.33
01-ai/Yi-34B-200K	50.00	42.86	60.00	26.67	66.67	55.56	37.50	85.71	33.33	50.00	33.33	25.00	12.50	22.22	40.00	30.00	42.86	33.33
01-ai/Yi-34B-Chat-4bits	56.25	57.14	53.33	26.67	33.33	55.56	50.00	71.43	22.22	37.50	16.67	25.00	0.00	22.22	30.00	30.00	71.43	11.11
01-ai/Yi-34B-Chat	68.75	57.14	60.00	46.67	16.67	77.78	50.00	71.43	22.22	43.75	16.67	25.00	12.50	22.22	30.00	40.00	42.86	11.11
CYFRAGOVPL/Llama-PLLuM-70B-base	50.00	71.43	66.67	73.33	66.67	77.78	62.50	100.00	77.78	68.75	50.00	25.00	12.50	44.44	20.00	60.00	42.86	33.33
CYFRAGOVPL/Llama-PLLuM-70B-chat	50.00	71.43	53.33	60.00	100.00	77.78	62.50	85.71	55.56	62.50	50.00	12.50	37.50	44.44	20.00	60.00	42.86	11.11
CYFRAGOVPL/Llama-PLLuM-8B-base	31.25	42.86	46.67	33.33	50.00	44.44	50.00	71.43	77.78	62.50	33.33	25.00	12.50	33.33	40.00	20.00	14.29	11.11
CYFRAGOVPL/Llama-PLLuM-8B-chat	31.25	28.57	33.33	40.00	50.00	44.44	50.00	57.14	55.56	37.50	0.00	37.50	12.50	22.22	40.00	10.00	14.29	0.00
CYFRAGOVPL/PLLuM-12B-base	37.50	57.14	46.67	26.67	33.33	66.67	50.00	71.43	55.56	37.50	33.33	50.00	0.00	11.11	40.00	10.00	42.86	11.11
CYFRAGOVPL/PLLuM-12B-chat	37.50	42.86	53.33	33.33	33.33	44.44	62.50	57.14	44.44	43.75	50.00	37.50	25.00	22.22	20.00	10.00	57.14	11.11
CYFRAGOVPL/PLLuM-12B-nc-base	18.75	42.86	53.33	46.67	66.67	44.44	75.00	57.14	55.56	50.00	0.00	50.00	37.50	11.11	30.00	20.00	42.86	44.44
CYFRAGOVPL/PLLuM-12B-nc-chat	18.75	42.86	46.67	46.67	50.00	66.67	37.50	57.14	44.44	43.75	0.00	12.50	0.00	22.22	30.00	20.00	42.86	11.11
CYFRAGOVPL/PLLuM-8x7B-base	31.25	57.14	60.00	46.67	50.00	55.56	62.50	71.43	44.44	62.50	16.67	50.00	12.50	44.44	30.00	40.00	14.29	44.44
CYFRAGOVPL/PLLuM-8x7B-chat	37.50	42.86	46.67	53.33	33.33	66.67	50.00	71.43	66.67	62.50	66.67	37.50	12.50	44.44	20.00	50.00	14.29	22.22
CYFRAGOVPL/PLLuM-8x7B-nc-base	43.75	71.43	53.33	73.33	66.67	77.78	62.50	71.43	55.56	50.00	33.33	12.50	0.00	55.56	40.00	50.00	42.86	44.44
CYFRAGOVPL/PLLuM-8x7B-nc-chat	43.75	42.86	46.67	66.67	33.33	77.78	62.50	85.71	44.44	56.25	66.67	37.50	25.00	55.56	20.00	40.00	42.86	33.33
CYFRAGOVPL/pllum-12b-nc-chat-250715	25.00	28.57	53.33	46.67	50.00	44.44	37.50	71.43	44.44	62.50	33.33	25.00	12.50	11.11	20.00	30.00	57.14	11.11
OPI-PG/Qra-13b	43.75	28.57	26.67	26.67	33.33	44.44	37.50	42.86	11.11	50.00	16.67	12.50	37.50	11.11	30.00	30.00	14.29	11.11
OPI-PG/Qra-1b	50.00	42.86	33.33	26.67	16.67	33.33	12.50	28.57	11.11	31.25	50.00	37.50	25.00	11.11	20.00	20.00	14.29	11.11
OPI-PG/Qra-7b	18.75	14.29	33.33	33.33	33.33	44.44	37.50	42.86	11.11	18.75	50.00	12.50	25.00	22.22	20.00	20.00	28.57	0.00
Qwen/Qwen2-1.5B	43.75	57.14	33.33	20.00	33.33	66.67	25.00	57.14	33.33	25.00	16.67	62.50	12.50	22.22	20.00	30.00	14.29	11.11
Qwen/Qwen2-7B	50.00	85.71	46.67	40.00	66.67	66.67	50.00	100.00	44.44	50.00	50.00	12.50	25.00	55.56	30.00	50.00	42.86	22.22
Qwen/Qwen2.5-14B	56.25	85.71	80.00	46.67	83.33	100.00	62.50	100.00	88.89	56.25	50.00	37.50	37.50	44.44	60.00	60.00	57.14	33.33
Qwen/Qwen2.5-32B	75.00	85.71	80.00	60.00	83.33	88.89	62.50	85.71	77.78	75.00	66.67	62.50	37.50	44.44	50.00	50.00	71.43	77.78
Qwen/Qwen2.5-72B-Instruct	68.75	85.71	86.67	73.33	100.00	88.89	87.50	100.00	88.89	81.25	100.00	62.50	62.50	33.33	70.00	40.00	57.14	55.56
Qwen/Qwen2.5-72B	75.00	85.71	86.67	80.00	83.33	88.89	87.50	100.00	88.89	81.25	83.33	62.50	62.50	44.44	80.00	50.00	85.71	66.67
Qwen/Qwen3-1.7B	25.00	57.14	53.33	40.00	66.67	55.56	37.50	71.43	44.44	43.75	50.00	25.00	25.00	22.22	10.00	20.00	42.86	44.44
Qwen/Qwen3-14B	62.50	100.00	66.67	73.33	83.33	88.89	62.50	100.00	66.67	75.00	83.33	25.00	37.50	22.22	70.00	30.00	42.86	33.33
Qwen/Qwen3-30B-A3B	50.00	85.71	86.67	73.33	83.33	88.89	75.00	100.00	66.67	75.00	100.00	75.00	62.50	77.78	70.00	30.00	42.86	33.33
Qwen/Qwen3-32B	81.25	100.00	100.00	80.00	66.67	100.00	75.00	100.00	88.89	68.75	100.00	62.50	75.00	66.67	70.00	60.00	85.71	66.67
Qwen/Qwen3-4B	50.00	71.43	66.67	66.67	83.33	77.78	75.00	71.43	66.67	43.75	66.67	37.50	37.50	44.44	50.00	30.00	14.29	66.67
Qwen/Qwen3-8B	68.75	100.00	66.67	66.67	100.00	77.78	75.00	100.00	77.78	56.25	66.67	50.00	25.00	44.44	30.00	40.00	28.57	44.44
Voicelab/trurl-2-13b-8bit	43.75	28.57	46.67	26.67	50.00	33.33	75.00	57.14	33.33	62.50	16.67	0.00	37.50	22.22	10.00	50.00	28.57	11.11
Voicelab/trurl-2-13b-academic	25.00	42.86	33.33	33.33	33.33	44.44	37.50	42.86	44.44	37.50	0.00	12.50	37.50	33.33	30.00	30.00	57.14	0.00
Voicelab/trurl-2-13b	31.25	28.57	46.67	33.33	50.00	33.33	75.00	57.14	33.33	62.50	0.00	0.00	25.00	22.22	30.00	40.00	28.57	11.11
Voicelab/trurl-2-7b-8bit	37.50	14.29	20.00	26.67	33.33	22.22	50.00	42.86	22.22	50.00	16.67	25.00	0.00	11.11	30.00	20.00	28.57	33.33
Voicelab/trurl-2-7b	31.25	14.29	20.00	40.00	33.33	22.22	50.00	57.14	22.22	56.25	33.33	25.00	12.50	11.11	30.00	10.00	28.57	22.22
allenai/OLMo-2-0325-32B-Instruct	62.50	42.86	46.67	40.00	50.00	44.44	50.00	85.71	77.78	43.75	33.33	25.00	0.00	55.56	20.00	40.00	14.29	22.22
allenai/OLMo-2-0325-32B	50.00	57.14	33.33	40.00	33.33	44.44	37.50	85.71	22.22	50.00	0.00	25.00	12.50	11.11	30.00	30.00	0.00	22.22
allenai/OLMo-2-1124-13B-Instruct	31.25	71.43	46.67	20.00	16.67	44.44	25.00	57.14	33.33	37.50	33.33	12.50	37.50	33.33	20.00	20.00	14.29	33.33
allenai/OLMo-2-1124-13B	43.75	42.86	60.00	26.67	16.67	44.44	25.00	71.43	22.22	25.00	16.67	12.50	25.00	22.22	40.00	40.00	42.86	33.33
allenai/OLMo-2-1124-7B-Instruct	43.75	14.29	46.67	33.33	16.67	22.22	12.50	57.14	33.33	37.50	33.33	12.50	50.00	11.11	20.00	20.00	14.29	11.11
allenai/OLMo-2-1124-7B	43.75	14.29	40.00	33.33	16.67	22.22	12.50	28.57	33.33	25.00	16.67	0.00	25.00	22.22	30.00	20.00	14.29	44.44
deepseek-ai/DeepSeek-R1-Distill-Qwen-14B	31.25	42.86	66.67	46.67	66.67	55.56	62.50	71.43	44.44	62.50	50.00	25.00	37.50	55.56	40.00	50.00	28.57	22.22
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	56.25	85.71	80.00	53.33	83.33	88.89	62.50	85.71	77.78	68.75	66.67	75.00	62.50	66.67	50.00	30.00	28.57	77.78
google/gemma-7b	37.50	57.14	60.00	46.67	50.00	66.67	50.00	85.71	44.44	62.50	0.00	25.00	12.50	55.56	20.00	30.00	14.29	33.33
lucyknada/microsoft_WizardLM-2-7B	25.00	57.14	60.00	40.00	33.33	55.56	50.00	57.14	33.33	37.50	0.00	25.00	12.50	22.22	20.00	30.00	57.14	33.33
meta-llama/Llama-2-13b-hf	37.50	42.86	33.33	40.00	16.67	44.44	37.50	42.86	11.11	37.50	16.67	12.50	25.00	11.11	30.00	10.00	14.29	22.22
meta-llama/Llama-2-70b-hf	31.25	42.86	40.00	40.00	50.00	55.56	37.50	85.71	66.67	50.00	33.33	25.00	62.50	33.33	40.00	30.00	42.86	22.22
meta-llama/Llama-2-7b-hf	12.50	14.29	26.67	20.00	33.33	33.33	62.50	42.86	22.22	18.75	16.67	12.50	25.00	11.11	10.00	20.00	14.29	22.22
meta-llama/Llama-3.1-70B	56.25	85.71	60.00	73.33	66.67	77.78	62.50	100.00	77.78	75.00	50.00	25.00	12.50	33.33	30.00	70.00	42.86	44.44
meta-llama/Llama-3.2-1B-Instruct	37.50	28.57	26.67	33.33	0.00	44.44	12.50	71.43	22.22	31.25	16.67	25.00	12.50	22.22	20.00	30.00	0.00	44.44
meta-llama/Llama-3.2-3B-Instruct	37.50	42.86	33.33	46.67	33.33	55.56	62.50	57.14	33.33	43.75	0.00	12.50	37.50	33.33	40.00	40.00	0.00	22.22
meta-llama/Llama-3.3-70B-Instruct	68.75	100.00	86.67	73.33	100.00	88.89	75.00	100.00	88.89	81.25	50.00	25.00	62.50	55.56	20.00	50.00	14.29	55.56
meta-llama/Meta-Llama-3-70B-Instruct	68.75	100.00	73.33	73.33	50.00	88.89	62.50	100.00	66.67	62.50	50.00	37.50	37.50	33.33	40.00	60.00	28.57	44.44
meta-llama/Meta-Llama-3-70B	62.50	57.14	66.67	73.33	83.33	88.89	62.50	100.00	77.78	75.00	50.00	12.50	12.50	44.44	20.00	40.00	28.57	44.44
meta-llama/Meta-Llama-3-8B-Instruct	43.75	28.57	53.33	26.67	66.67	44.44	62.50	85.71	44.44	50.00	33.33	0.00	0.00	44.44	30.00	10.00	42.86	33.33
meta-llama/Meta-Llama-3-8B	43.75	42.86	33.33	46.67	33.33	55.56	37.50	85.71	33.33	43.75	33.33	25.00	25.00	11.11	50.00	30.00	28.57	33.33
meta-llama/Meta-Llama-3.1-70B-Instruct	68.75	100.00	73.33	73.33	100.00	88.89	62.50	100.00	77.78	62.50	50.00	25.00	37.50	55.56	40.00	50.00	28.57	44.44
meta-llama/Meta-Llama-3.1-8B-Instruct	25.00	71.43	33.33	53.33	83.33	66.67	62.50	85.71	44.44	50.00	0.00	25.00	25.00	33.33	40.00	40.00	28.57	33.33
meta-llama/Meta-Llama-3.1-8B	25.00	57.14	40.00	33.33	50.00	55.56	37.50	71.43	77.78	31.25	33.33	12.50	25.00	22.22	30.00	0.00	0.00	22.22
microsoft/Phi-3-mini-4k-instruct	37.50	42.86	20.00	60.00	16.67	22.22	37.50	57.14	11.11	43.75	16.67	0.00	37.50	44.44	30.00	40.00	14.29	33.33
microsoft/phi-1	25.00	42.86	26.67	33.33	16.67	44.44	25.00	57.14	11.11	31.25	16.67	12.50	50.00	0.00	20.00	30.00	28.57	22.22
microsoft/phi-1_5	25.00	28.57	20.00	20.00	33.33	33.33	12.50	28.57	33.33	6.25	33.33	37.50	12.50	11.11	20.00	20.00	42.86	0.00
microsoft/phi-2	25.00	42.86	20.00	13.33	50.00	33.33	12.50	28.57	33.33	6.25	33.33	25.00	12.50	22.22	20.00	10.00	42.86	11.11
microsoft/phi-4	50.00	71.43	60.00	66.67	83.33	55.56	50.00	85.71	55.56	68.75	66.67	50.00	50.00	33.33	50.00	50.00	57.14	44.44
mistralai/Ministral-8B-Instruct-2410	50.00	57.14	53.33	40.00	33.33	44.44	62.50	71.43	33.33	56.25	16.67	37.50	25.00	22.22	30.00	50.00	28.57	22.22
mistralai/Mistral-7B-Instruct-v0.1	12.50	28.57	26.67	40.00	33.33	55.56	37.50	57.14	0.00	18.75	16.67	25.00	0.00	33.33	20.00	30.00	28.57	33.33
mistralai/Mistral-7B-Instruct-v0.2	37.50	57.14	60.00	40.00	50.00	66.67	50.00	71.43	66.67	50.00	16.67	37.50	50.00	33.33	30.00	20.00	42.86	33.33
mistralai/Mistral-7B-Instruct-v0.3	31.25	57.14	60.00	46.67	33.33	77.78	62.50	57.14	55.56	43.75	16.67	25.00	37.50	55.56	20.00	30.00	57.14	33.33
mistralai/Mistral-7B-v0.1	43.75	28.57	53.33	46.67	16.67	77.78	87.50	57.14	33.33	31.25	16.67	0.00	25.00	22.22	30.00	30.00	28.57	22.22
mistralai/Mistral-7B-v0.3	43.75	28.57	40.00	26.67	33.33	55.56	62.50	28.57	33.33	37.50	16.67	0.00	37.50	33.33	50.00	20.00	14.29	44.44
mistralai/Mistral-Large-Instruct-2407	68.75	100.00	73.33	73.33	83.33	88.89	87.50	100.00	88.89	81.25	66.67	75.00	50.00	55.56	40.00	50.00	85.71	22.22
mistralai/Mistral-Large-Instruct-2411	68.75	100.00	80.00	73.33	50.00	88.89	87.50	100.00	88.89	81.25	83.33	75.00	37.50	55.56	40.00	50.00	85.71	33.33
mistralai/Mistral-Nemo-Base-2407	50.00	14.29	40.00	20.00	33.33	55.56	37.50	71.43	33.33	43.75	33.33	25.00	37.50	22.22	40.00	30.00	28.57	11.11
mistralai/Mistral-Nemo-Instruct-2407	37.50	28.57	66.67	40.00	50.00	44.44	62.50	57.14	33.33	43.75	50.00	12.50	37.50	55.56	50.00	30.00	28.57	33.33
mistralai/Mixtral-8x22B-Instruct-v0.1	43.75	85.71	60.00	73.33	83.33	44.44	50.00	85.71	33.33	68.75	83.33	12.50	12.50	55.56	40.00	40.00	28.57	33.33
mistralai/Mixtral-8x7B-Instruct-v0.1	31.25	57.14	46.67	53.33	16.67	66.67	50.00	57.14	44.44	56.25	16.67	25.00	12.50	66.67	40.00	60.00	28.57	33.33
mistralai/Mixtral-8x7B-v0.1	31.25	57.14	46.67	60.00	66.67	66.67	62.50	57.14	44.44	43.75	0.00	37.50	12.50	44.44	40.00	60.00	42.86	44.44
sdadas/polish-gpt2-large	18.75	14.29	13.33	13.33	33.33	33.33	0.00	14.29	11.11	12.50	50.00	12.50	25.00	22.22	10.00	20.00	14.29	22.22
sdadas/polish-gpt2-medium	31.25	28.57	20.00	13.33	33.33	44.44	37.50	14.29	33.33	18.75	33.33	25.00	25.00	11.11	20.00	30.00	28.57	0.00
sdadas/polish-gpt2-small	31.25	28.57	20.00	20.00	16.67	33.33	12.50	28.57	33.33	6.25	33.33	37.50	25.00	0.00	10.00	20.00	57.14	0.00
sdadas/polish-gpt2-xl	31.25	28.57	13.33	20.00	16.67	33.33	12.50	28.57	22.22	6.25	33.33	25.00	25.00	0.00	10.00	20.00	42.86	0.00
speakleash/Bielik-1.5B-v3.0-Instruct	31.25	14.29	53.33	26.67	50.00	66.67	50.00	71.43	55.56	50.00	0.00	0.00	37.50	33.33	40.00	30.00	14.29	33.33
speakleash/Bielik-11B-v2.0-Instruct	56.25	57.14	60.00	40.00	50.00	44.44	62.50	57.14	66.67	56.25	16.67	37.50	12.50	44.44	30.00	40.00	28.57	11.11
speakleash/Bielik-11B-v2.1-Instruct	56.25	42.86	53.33	46.67	66.67	55.56	62.50	85.71	55.56	62.50	0.00	37.50	50.00	55.56	30.00	30.00	28.57	33.33
speakleash/Bielik-11B-v2.2-Instruct	56.25	42.86	53.33	46.67	66.67	55.56	62.50	85.71	55.56	56.25	16.67	37.50	37.50	55.56	30.00	30.00	28.57	44.44
speakleash/Bielik-11B-v2.3-Instruct_newer	56.25	42.86	53.33	53.33	83.33	44.44	62.50	85.71	66.67	68.75	16.67	50.00	25.00	44.44	30.00	30.00	28.57	33.33
speakleash/Bielik-11B-v2.5-Instruct	37.50	57.14	53.33	60.00	66.67	77.78	62.50	100.00	66.67	43.75	16.67	12.50	25.00	55.56	40.00	40.00	28.57	44.44
speakleash/Bielik-11B-v2.6-Instruct	37.50	57.14	66.67	60.00	50.00	77.78	62.50	100.00	66.67	43.75	16.67	12.50	25.00	44.44	40.00	40.00	28.57	44.44
speakleash/Bielik-11B-v2	37.50	42.86	66.67	66.67	50.00	66.67	50.00	100.00	77.78	56.25	33.33	25.00	37.50	33.33	10.00	50.00	42.86	44.44
speakleash/Bielik-4.5B-v3.0-Instruct	31.25	42.86	53.33	40.00	66.67	55.56	62.50	85.71	66.67	50.00	33.33	37.50	25.00	33.33	10.00	60.00	0.00	11.11
speakleash/Bielik-7B-Instruct-v0.1	31.25	28.57	13.33	13.33	33.33	55.56	50.00	57.14	44.44	43.75	16.67	0.00	62.50	0.00	20.00	20.00	14.29	44.44
speakleash/Bielik-7B-v0.1	37.50	42.86	33.33	26.67	33.33	44.44	62.50	71.43	22.22	50.00	16.67	12.50	37.50	22.22	30.00	30.00	14.29	22.22
utter-project/EuroLLM-1.7B-Instruct	37.50	0.00	40.00	26.67	0.00	33.33	25.00	28.57	33.33	25.00	33.33	12.50	25.00	55.56	30.00	10.00	28.57	33.33
utter-project/EuroLLM-9B-Instruct	25.00	42.86	46.67	33.33	50.00	77.78	62.50	71.43	66.67	43.75	33.33	37.50	25.00	11.11	40.00	20.00	0.00	33.33

8th-Grade Exam

The 8th-grade exam (egzamin ósmoklasisty) is a standardized assessment taken by Polish students at the end of their primary education (8th grade).

Subjects Covered

This exam mainly includes tests in:

Polish Language
Mathematics
Foreign Language

Purpose and Role

Secondary Education Foundation: Provides a foundation for secondary education placements, guiding students to the next stage of their academic journey.
Student Assessment: Evaluates core competencies, ensuring readiness for high school-level studies.

Model	2019	2020	2021	2022	2023
speakleash/Bielik-11B-v2.3-Instruct_newer	50	41.67	50	36.36	71.43

Model	2019	2020	2021	2022	2023
01-ai/Yi-1.5-34B-Chat	50.00	50.00	50.00	36.36	71.43
01-ai/Yi-1.5-34B	50.00	41.67	40.00	54.55	42.86
01-ai/Yi-1.5-9B-Chat	50.00	25.00	30.00	36.36	42.86
01-ai/Yi-1.5-9B	50.00	16.67	20.00	45.45	28.57
01-ai/Yi-34B-200K	40.00	16.67	50.00	54.55	42.86
01-ai/Yi-34B-Chat-4bits	20.00	41.67	50.00	36.36	42.86
01-ai/Yi-34B-Chat	20.00	50.00	40.00	27.27	14.29
CYFRAGOVPL/Llama-PLLuM-70B-base	50.00	33.33	40.00	27.27	42.86
CYFRAGOVPL/Llama-PLLuM-70B-chat	70.00	25.00	40.00	54.55	42.86
CYFRAGOVPL/Llama-PLLuM-8B-base	40.00	8.33	50.00	27.27	14.29
CYFRAGOVPL/Llama-PLLuM-8B-chat	50.00	16.67	50.00	27.27	42.86
CYFRAGOVPL/PLLuM-12B-base	40.00	16.67	50.00	54.55	42.86
CYFRAGOVPL/PLLuM-12B-chat	60.00	25.00	40.00	27.27	28.57
CYFRAGOVPL/PLLuM-12B-nc-base	30.00	25.00	40.00	27.27	57.14
CYFRAGOVPL/PLLuM-12B-nc-chat	50.00	25.00	60.00	27.27	57.14
CYFRAGOVPL/PLLuM-8x7B-base	30.00	25.00	50.00	18.18	28.57
CYFRAGOVPL/PLLuM-8x7B-chat	40.00	16.67	40.00	36.36	42.86
CYFRAGOVPL/PLLuM-8x7B-nc-base	40.00	33.33	40.00	27.27	14.29
CYFRAGOVPL/PLLuM-8x7B-nc-chat	50.00	33.33	40.00	27.27	42.86
CYFRAGOVPL/pllum-12b-nc-chat-250715	50.00	16.67	50.00	36.36	57.14
OPI-PG/Qra-13b	20.00	0.00	50.00	45.45	42.86
OPI-PG/Qra-1b	30.00	16.67	20.00	27.27	57.14
OPI-PG/Qra-7b	10.00	16.67	30.00	27.27	14.29
Qwen/Qwen2-1.5B	20.00	16.67	20.00	18.18	28.57
Qwen/Qwen2-7B	60.00	33.33	40.00	18.18	42.86
Qwen/Qwen2.5-14B	70.00	41.67	60.00	27.27	28.57
Qwen/Qwen2.5-32B	70.00	50.00	50.00	36.36	71.43
Qwen/Qwen2.5-72B-Instruct	80.00	50.00	50.00	63.64	100.00
Qwen/Qwen2.5-72B	80.00	41.67	50.00	54.55	100.00
Qwen/Qwen3-1.7B	20.00	25.00	30.00	45.45	14.29
Qwen/Qwen3-14B	70.00	41.67	60.00	36.36	71.43
Qwen/Qwen3-30B-A3B	50.00	41.67	60.00	36.36	71.43
Qwen/Qwen3-32B	70.00	50.00	50.00	72.73	85.71
Qwen/Qwen3-4B	30.00	58.33	30.00	27.27	42.86
Qwen/Qwen3-8B	60.00	58.33	60.00	18.18	57.14
Voicelab/trurl-2-13b-8bit	30.00	33.33	40.00	54.55	14.29
Voicelab/trurl-2-13b-academic	30.00	33.33	40.00	36.36	14.29
Voicelab/trurl-2-13b	30.00	25.00	50.00	54.55	14.29
Voicelab/trurl-2-7b-8bit	20.00	25.00	10.00	9.09	28.57
Voicelab/trurl-2-7b	40.00	25.00	10.00	9.09	28.57
allenai/OLMo-2-0325-32B-Instruct	50.00	41.67	40.00	9.09	57.14
allenai/OLMo-2-0325-32B	40.00	8.33	50.00	27.27	42.86
allenai/OLMo-2-1124-13B-Instruct	40.00	25.00	30.00	9.09	57.14
allenai/OLMo-2-1124-13B	60.00	25.00	20.00	54.55	42.86
allenai/OLMo-2-1124-7B-Instruct	40.00	33.33	50.00	45.45	28.57
allenai/OLMo-2-1124-7B	50.00	33.33	30.00	45.45	28.57
deepseek-ai/DeepSeek-R1-Distill-Qwen-14B	20.00	50.00	30.00	54.55	57.14
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	60.00	41.67	50.00	45.45	71.43
google/gemma-7b	50.00	41.67	50.00	27.27	14.29
lucyknada/microsoft_WizardLM-2-7B	50.00	25.00	20.00	27.27	42.86
meta-llama/Llama-2-13b-hf	20.00	25.00	40.00	36.36	42.86
meta-llama/Llama-2-70b-hf	70.00	8.33	40.00	18.18	28.57
meta-llama/Llama-2-7b-hf	20.00	16.67	20.00	18.18	14.29
meta-llama/Llama-3.1-70B	50.00	33.33	40.00	18.18	42.86
meta-llama/Llama-3.2-1B-Instruct	30.00	16.67	10.00	27.27	28.57
meta-llama/Llama-3.2-3B-Instruct	50.00	25.00	40.00	36.36	28.57
meta-llama/Llama-3.3-70B-Instruct	90.00	33.33	40.00	27.27	42.86
meta-llama/Meta-Llama-3-70B-Instruct	70.00	41.67	40.00	54.55	71.43
meta-llama/Meta-Llama-3-70B	50.00	33.33	40.00	18.18	42.86
meta-llama/Meta-Llama-3-8B-Instruct	60.00	16.67	40.00	45.45	28.57
meta-llama/Meta-Llama-3-8B	60.00	16.67	50.00	36.36	57.14
meta-llama/Meta-Llama-3.1-70B-Instruct	90.00	41.67	40.00	54.55	42.86
meta-llama/Meta-Llama-3.1-8B-Instruct	50.00	25.00	50.00	45.45	28.57
meta-llama/Meta-Llama-3.1-8B	30.00	25.00	60.00	27.27	28.57
microsoft/Phi-3-mini-4k-instruct	30.00	25.00	40.00	18.18	14.29
microsoft/phi-1	30.00	16.67	30.00	45.45	14.29
microsoft/phi-1_5	30.00	16.67	30.00	36.36	28.57
microsoft/phi-2	20.00	16.67	30.00	18.18	14.29
microsoft/phi-4	60.00	25.00	70.00	36.36	57.14
mistralai/Ministral-8B-Instruct-2410	60.00	16.67	50.00	9.09	28.57
mistralai/Mistral-7B-Instruct-v0.1	30.00	33.33	20.00	36.36	14.29
mistralai/Mistral-7B-Instruct-v0.2	40.00	25.00	30.00	27.27	42.86
mistralai/Mistral-7B-Instruct-v0.3	40.00	8.33	40.00	27.27	14.29
mistralai/Mistral-7B-v0.1	50.00	8.33	40.00	18.18	14.29
mistralai/Mistral-7B-v0.3	30.00	25.00	30.00	18.18	14.29
mistralai/Mistral-Large-Instruct-2407	60.00	41.67	50.00	36.36	71.43
mistralai/Mistral-Large-Instruct-2411	70.00	41.67	60.00	9.09	71.43
mistralai/Mistral-Nemo-Base-2407	40.00	16.67	40.00	27.27	42.86
mistralai/Mistral-Nemo-Instruct-2407	40.00	25.00	60.00	36.36	28.57
mistralai/Mixtral-8x22B-Instruct-v0.1	70.00	50.00	60.00	36.36	57.14
mistralai/Mixtral-8x7B-Instruct-v0.1	20.00	33.33	40.00	9.09	28.57
mistralai/Mixtral-8x7B-v0.1	50.00	8.33	40.00	9.09	28.57
sdadas/polish-gpt2-large	30.00	8.33	40.00	45.45	14.29
sdadas/polish-gpt2-medium	10.00	16.67	20.00	54.55	28.57
sdadas/polish-gpt2-small	10.00	16.67	30.00	27.27	14.29
sdadas/polish-gpt2-xl	10.00	16.67	30.00	27.27	14.29
speakleash/Bielik-1.5B-v3.0-Instruct	30.00	50.00	30.00	18.18	28.57
speakleash/Bielik-11B-v2.0-Instruct	70.00	25.00	60.00	36.36	42.86
speakleash/Bielik-11B-v2.1-Instruct	80.00	33.33	40.00	18.18	57.14
speakleash/Bielik-11B-v2.2-Instruct	80.00	25.00	50.00	36.36	57.14
speakleash/Bielik-11B-v2.3-Instruct_newer	80.00	25.00	50.00	27.27	57.14
speakleash/Bielik-11B-v2.5-Instruct	60.00	58.33	40.00	45.45	42.86
speakleash/Bielik-11B-v2.6-Instruct	60.00	58.33	40.00	36.36	57.14
speakleash/Bielik-11B-v2	40.00	16.67	40.00	36.36	28.57
speakleash/Bielik-4.5B-v3.0-Instruct	40.00	33.33	60.00	36.36	57.14
speakleash/Bielik-7B-Instruct-v0.1	60.00	41.67	30.00	18.18	28.57
speakleash/Bielik-7B-v0.1	50.00	16.67	20.00	36.36	28.57
utter-project/EuroLLM-1.7B-Instruct	60.00	25.00	20.00	27.27	14.29
utter-project/EuroLLM-9B-Instruct	50.00	8.33	30.00	36.36	28.57

High School Exam

The High School Exam (matura) is one of the most critical exams in Poland. This exam serves as a prerequisite for higher education and is a significant milestone in students' academic journeys.

Mandatory Subjects

The matura exam includes essential subjects:

Polish Language
Mathematics
Foreign Language

Additional Subjects

Students may also choose additional subjects based on:

Areas of Interest: Students select subjects aligned with their strengths or future studies.
University Requirements: Specific subjects may be necessary for admission to certain university programs.

Minimum points required

To pass the High School Exam, students needed to achieve a minimum score of 30% in mandatory subjects. Additional subjects do not require a minimum score to pass, but the results significantly impact the points needed for university admission.

Model	2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023
speakleash/Bielik-11B-v2.3-Instruct_newer	20	42.86	60	62.5	66.67	33.33	71.43	100	55.56	62.5	12.5	100	100	38.71	51.61	32.14	41.18	25.93	39.29	32	38.71	48.89

Professional Exam

The Professional Exam (egzamin zawodowy) is tailored for students pursuing vocational education in Poland. This exam is typically taken after completing vocational training programs, which may range from:

Technical High Schools
Post-secondary Vocational Courses

Purpose and Importance

Certification: A successful score on this exam certifies a student’s qualifications in a specific trade or profession.
Career Readiness: Enables students to enter the workforce directly.
Further Education: Offers a foundation for pursuing specialized training.

Model	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
speakleash/Bielik-11B-v2.3-Instruct_newer	38.29	39.47	41.06	36.54	41.01	40.99	42.58	42.06	42.76	42.07	39.51	41.43

Model	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022	2023	2024
01-ai/Yi-1.5-34B-Chat	38.29	39.47	41.06	36.54	41.01	40.99	42.58	42.06	42.76	42.07	39.51	41.43
01-ai/Yi-1.5-34B	36.86	40.07	35.43	34.62	41.34	41.50	42.26	43.06	40.90	40.21	41.35	39.05
01-ai/Yi-1.5-9B-Chat	37.27	37.32	39.40	31.82	38.12	37.09	38.67	39.56	36.31	36.89	35.53	40.48
01-ai/Yi-1.5-9B	35.44	33.10	28.81	33.04	38.19	37.57	38.55	38.21	36.77	37.88	36.29	36.43
01-ai/Yi-34B-200K	32.79	34.57	35.10	29.72	36.98	37.70	38.32	40.20	38.08	38.87	35.22	39.52
01-ai/Yi-34B-Chat-4bits	35.44	38.95	37.75	33.39	38.86	42.15	42.50	40.55	41.33	40.09	39.20	40.95
01-ai/Yi-34B-Chat	35.64	39.04	39.40	35.84	40.54	42.37	43.08	42.80	42.37	41.84	40.58	40.71
CYFRAGOVPL/Llama-PLLuM-70B-base	57.23	64.23	63.58	61.54	64.36	67.01	66.51	66.65	62.65	66.26	64.47	61.67
CYFRAGOVPL/Llama-PLLuM-70B-chat	57.84	64.75	62.91	61.36	64.70	65.46	66.98	65.62	63.85	64.34	66.77	62.86
CYFRAGOVPL/Llama-PLLuM-8B-base	38.29	44.80	45.36	44.06	46.17	47.86	48.91	47.20	47.39	45.69	44.10	45.48
CYFRAGOVPL/Llama-PLLuM-8B-chat	44.40	46.78	49.01	44.58	46.71	49.72	49.79	49.20	48.36	46.56	46.25	44.76
CYFRAGOVPL/PLLuM-12B-base	39.51	43.25	41.39	43.18	45.64	48.38	49.12	48.26	46.85	45.28	47.78	47.14
CYFRAGOVPL/PLLuM-12B-chat	45.21	47.81	43.71	45.98	48.39	51.53	51.91	51.00	50.79	50.12	48.70	48.81
CYFRAGOVPL/PLLuM-12B-nc-base	47.45	49.10	48.34	47.03	50.00	52.18	53.86	51.16	51.02	49.59	47.32	54.76
CYFRAGOVPL/PLLuM-12B-nc-chat	48.68	52.54	50.33	48.43	52.95	54.95	55.30	55.98	53.73	53.73	53.14	52.62
CYFRAGOVPL/PLLuM-8x7B-base	48.07	50.30	51.32	46.33	51.48	53.65	54.83	53.76	52.68	51.92	52.83	49.76
CYFRAGOVPL/PLLuM-8x7B-chat	45.82	50.21	50.66	45.98	51.54	55.04	54.36	55.82	53.53	52.86	50.54	55.71
CYFRAGOVPL/PLLuM-8x7B-nc-base	52.75	55.80	56.62	51.05	58.39	60.48	59.92	59.67	57.47	59.44	56.36	54.76
CYFRAGOVPL/PLLuM-8x7B-nc-chat	54.79	59.42	56.62	52.97	59.66	62.99	62.74	63.43	59.79	62.53	59.57	60.95
CYFRAGOVPL/pllum-12b-nc-chat-250715	48.27	55.12	48.68	51.75	54.43	56.12	56.09	56.07	54.69	54.78	52.37	53.57
OPI-PG/Qra-13b	30.35	31.30	31.13	33.39	35.57	35.71	36.90	35.80	35.15	35.55	30.63	33.81
OPI-PG/Qra-1b	23.63	24.16	24.50	26.05	25.91	25.25	26.43	25.51	26.07	24.24	23.28	27.38
OPI-PG/Qra-7b	27.90	29.75	23.18	27.62	30.13	27.84	28.49	31.68	26.84	31.35	27.87	29.76
Qwen/Qwen2-1.5B	30.75	35.68	41.72	32.34	34.23	34.11	35.31	34.38	33.49	33.16	33.54	35.95
Qwen/Qwen2-7B	37.47	42.48	40.40	41.08	45.03	46.43	47.03	47.62	45.62	46.04	43.34	46.43
Qwen/Qwen2.5-14B	51.53	55.80	52.98	49.65	54.56	57.07	56.36	55.49	55.20	54.95	55.44	51.19
Qwen/Qwen2.5-32B	56.62	58.13	59.27	56.64	61.74	62.65	62.24	61.25	60.53	61.01	58.96	57.62
Qwen/Qwen2.5-72B-Instruct	63.14	65.78	65.23	66.08	70.47	70.60	70.07	70.34	68.37	68.01	68.91	64.05
Qwen/Qwen2.5-72B	62.53	64.66	65.23	65.38	69.87	69.48	68.72	70.47	68.10	68.01	69.22	65.24
Qwen/Qwen3-1.7B	33.60	40.07	37.42	38.81	39.33	38.91	40.29	40.52	41.79	40.73	36.14	40.24
Qwen/Qwen3-14B	57.84	58.04	61.59	55.07	62.21	62.43	62.01	62.53	59.68	60.66	60.34	57.62
Qwen/Qwen3-30B-A3B	62.53	60.88	62.91	60.31	63.69	64.55	64.92	64.49	63.46	61.42	62.94	58.33
Qwen/Qwen3-32B	61.30	64.66	65.89	65.03	65.44	67.66	68.63	67.80	65.74	66.67	64.17	63.81
Qwen/Qwen3-4B	48.68	48.93	47.68	48.25	51.01	48.25	51.53	52.57	49.21	50.17	48.39	44.52
Qwen/Qwen3-8B	50.10	52.54	60.60	50.87	53.02	56.38	56.15	57.29	54.92	54.14	54.52	51.43
Voicelab/trurl-2-13b-8bit	35.03	37.66	43.05	37.76	40.34	40.86	42.11	40.87	40.67	40.73	39.20	37.38
Voicelab/trurl-2-13b-academic	31.36	33.79	33.77	33.57	34.50	35.58	35.14	35.76	35.50	35.43	34.92	34.76
Voicelab/trurl-2-13b	35.03	39.81	43.71	34.79	39.26	41.16	41.55	41.29	41.02	40.15	40.12	37.38
Voicelab/trurl-2-7b-8bit	32.38	30.70	32.45	33.92	32.15	31.86	32.90	31.46	31.52	31.64	31.85	32.14
Voicelab/trurl-2-7b	34.22	31.04	30.46	34.62	32.42	32.08	33.78	32.62	31.48	31.70	30.78	32.62
allenai/OLMo-2-0325-32B-Instruct	42.57	40.58	45.36	40.73	44.50	44.88	45.53	46.40	43.41	44.76	42.73	41.67
allenai/OLMo-2-0325-32B	36.66	39.12	43.05	38.81	43.09	42.41	43.91	43.93	41.25	41.96	36.75	40.24
allenai/OLMo-2-1124-13B-Instruct	29.74	31.81	29.14	32.34	33.83	35.11	33.43	33.64	33.22	33.74	33.54	32.38
allenai/OLMo-2-1124-13B	27.90	28.55	28.81	30.59	32.75	33.64	34.37	33.16	32.48	32.93	33.23	32.86
allenai/OLMo-2-1124-7B-Instruct	30.75	29.75	29.47	29.37	30.60	32.34	32.70	31.11	32.72	31.93	29.40	30.48
allenai/OLMo-2-1124-7B	31.16	29.92	30.79	28.32	31.41	31.26	32.05	31.52	31.05	31.12	31.39	30.24
deepseek-ai/DeepSeek-R1-Distill-Qwen-14B	42.77	43.85	45.70	43.53	45.57	48.08	48.56	47.27	46.43	45.92	46.55	45.71
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	52.95	54.69	53.31	52.80	54.50	56.90	58.00	55.62	55.74	53.96	52.68	53.10
google/gemma-7b	45.01	44.54	44.04	45.80	47.18	47.04	49.32	47.49	46.89	45.86	48.85	44.29
lucyknada/microsoft_WizardLM-2-7B	35.03	36.54	34.77	34.09	39.87	37.22	39.11	40.78	37.58	40.44	35.38	35.95
meta-llama/Llama-2-13b-hf	29.33	30.87	33.44	32.87	31.34	35.19	36.43	34.58	34.26	33.92	29.86	35.00
meta-llama/Llama-2-70b-hf	38.90	44.80	47.02	41.78	46.38	46.22	48.06	47.62	47.78	48.14	43.80	44.52
meta-llama/Llama-2-7b-hf	27.90	29.06	28.15	27.80	26.38	28.62	27.31	28.66	28.31	28.38	28.64	27.14
meta-llama/Llama-3.1-70B	56.21	61.65	64.90	58.22	63.69	65.89	64.45	64.30	63.07	63.58	63.71	58.57
meta-llama/Llama-3.2-1B-Instruct	29.33	30.87	33.77	30.42	31.81	32.38	31.87	31.43	32.29	31.29	31.24	33.81
meta-llama/Llama-3.2-3B-Instruct	36.86	38.44	39.74	34.62	38.52	41.25	40.88	41.61	39.86	40.56	38.59	42.62
meta-llama/Llama-3.3-70B-Instruct	61.51	64.49	64.57	63.11	67.45	69.48	68.36	68.03	66.94	68.12	67.23	64.05
meta-llama/Meta-Llama-3-70B-Instruct	57.23	62.08	62.91	58.92	63.42	66.36	66.98	65.26	63.89	63.46	63.09	58.81
meta-llama/Meta-Llama-3-70B	57.23	59.50	60.60	59.09	62.62	64.76	63.77	63.50	62.50	63.58	60.18	57.86
meta-llama/Meta-Llama-3-8B-Instruct	41.34	43.08	40.07	40.56	42.75	46.04	45.38	47.24	46.20	45.45	42.11	47.38
meta-llama/Meta-Llama-3-8B	39.51	38.87	34.44	37.41	41.81	41.55	43.73	42.38	42.68	40.56	37.37	41.43
meta-llama/Meta-Llama-3.1-70B-Instruct	62.12	64.75	61.92	64.86	66.51	68.66	68.45	67.71	66.40	66.78	63.40	63.10
meta-llama/Meta-Llama-3.1-8B-Instruct	46.44	45.23	51.66	42.48	45.91	49.24	49.00	49.13	47.43	46.85	47.47	45.24
meta-llama/Meta-Llama-3.1-8B	41.14	42.05	39.74	39.16	44.09	43.93	46.85	46.18	44.53	44.35	42.73	42.86
microsoft/Phi-3-mini-4k-instruct	27.90	30.78	32.78	27.45	33.89	33.72	34.43	34.00	34.11	35.08	32.16	34.29
microsoft/phi-1	25.05	23.65	23.51	24.65	25.70	26.63	26.02	26.57	24.91	25.87	25.73	25.95
microsoft/phi-1_5	20.77	25.54	22.19	26.40	25.10	23.13	23.28	25.51	23.64	24.30	25.11	25.24
microsoft/phi-2	21.38	26.66	22.85	25.52	26.17	25.51	24.72	26.06	25.03	27.10	27.57	28.10
microsoft/phi-4	46.84	48.58	44.70	44.76	51.21	53.78	53.33	53.53	50.79	51.63	50.38	52.62
mistralai/Ministral-8B-Instruct-2410	34.83	38.52	42.72	38.29	40.20	43.58	45.32	43.70	42.84	42.54	39.36	42.62
mistralai/Mistral-7B-Instruct-v0.1	36.05	34.65	38.08	33.92	35.57	36.32	37.43	36.60	36.46	36.25	33.69	36.43
mistralai/Mistral-7B-Instruct-v0.2	36.66	40.67	37.75	41.26	40.67	41.20	42.67	42.10	39.94	41.14	35.53	41.19
mistralai/Mistral-7B-Instruct-v0.3	40.33	39.72	41.39	40.38	41.07	42.24	43.50	43.93	40.32	43.01	37.98	40.24
mistralai/Mistral-7B-v0.1	39.51	35.94	39.07	36.19	38.32	38.87	39.02	38.40	37.39	38.11	33.69	35.71
mistralai/Mistral-7B-v0.3	35.44	35.77	39.74	33.57	37.25	37.35	38.08	38.85	37.08	37.30	34.76	35.48
mistralai/Mistral-Large-Instruct-2407	59.88	63.37	63.58	61.36	67.32	70.04	69.31	68.16	66.67	67.60	67.99	60.95
mistralai/Mistral-Large-Instruct-2411	59.27	63.63	64.24	60.84	66.11	68.96	68.33	68.86	65.31	67.48	67.38	60.95
mistralai/Mistral-Nemo-Base-2407	35.64	40.07	38.08	38.46	43.69	43.62	44.41	44.15	41.75	41.67	38.13	41.19
mistralai/Mistral-Nemo-Instruct-2407	40.73	44.20	45.36	42.31	45.37	47.60	48.12	47.01	45.15	44.00	42.42	44.29
mistralai/Mixtral-8x22B-Instruct-v0.1	51.32	55.55	54.97	51.40	54.30	59.92	61.21	59.77	57.51	59.50	59.42	58.33
mistralai/Mixtral-8x7B-Instruct-v0.1	46.44	46.09	46.69	47.20	48.39	51.32	51.53	51.41	49.86	48.25	48.85	48.10
mistralai/Mixtral-8x7B-v0.1	44.81	45.74	47.68	45.28	47.85	48.90	48.88	49.23	48.59	47.32	45.79	43.81
sdadas/polish-gpt2-large	22.20	24.59	25.17	27.45	25.91	23.65	23.72	25.67	25.53	25.12	26.65	25.00
sdadas/polish-gpt2-medium	26.48	22.53	23.84	24.48	25.44	24.43	24.13	24.97	24.53	23.66	25.73	21.43
sdadas/polish-gpt2-small	20.57	24.85	22.19	25.35	24.23	23.00	23.63	25.35	24.10	24.24	25.73	25.95
sdadas/polish-gpt2-xl	20.98	25.19	21.52	24.65	24.77	23.39	24.19	24.42	23.29	23.19	25.11	26.19
speakleash/Bielik-1.5B-v3.0-Instruct	43.38	45.06	43.38	46.50	45.84	47.56	49.06	48.36	46.85	46.74	46.55	43.81
speakleash/Bielik-11B-v2.0-Instruct	50.31	54.34	54.97	54.55	55.57	56.38	57.09	58.26	56.39	53.26	54.82	55.95
speakleash/Bielik-11B-v2.1-Instruct	52.14	54.00	54.97	57.69	57.32	57.50	59.45	60.51	58.36	56.35	58.19	56.67
speakleash/Bielik-11B-v2.2-Instruct	51.93	53.83	52.65	57.69	57.11	57.80	58.18	60.51	58.67	56.53	57.27	57.86
speakleash/Bielik-11B-v2.3-Instruct_newer	51.93	54.08	53.64	56.82	56.51	58.06	58.56	60.12	58.94	56.53	58.81	56.90
speakleash/Bielik-11B-v2.5-Instruct	52.75	55.89	54.97	58.22	58.52	59.40	61.18	61.28	60.76	56.12	59.42	57.86
speakleash/Bielik-11B-v2.6-Instruct	52.95	55.63	56.62	56.64	58.52	60.05	62.18	61.60	60.49	56.76	58.81	58.10
speakleash/Bielik-11B-v2	51.53	54.60	54.64	52.62	55.44	56.12	57.45	56.59	55.23	54.02	55.59	51.43
speakleash/Bielik-4.5B-v3.0-Instruct	46.44	54.00	55.30	51.57	52.75	57.59	55.94	56.81	54.62	54.08	58.35	51.90
speakleash/Bielik-7B-Instruct-v0.1	37.27	36.63	38.41	36.71	40.87	41.72	42.17	42.00	41.91	41.84	39.82	41.19
speakleash/Bielik-7B-v0.1	35.85	38.01	35.10	37.94	37.58	39.52	40.73	40.49	39.44	40.50	36.60	39.29
utter-project/EuroLLM-1.7B-Instruct	24.44	25.80	25.83	24.65	27.32	26.03	25.25	24.97	23.99	28.03	23.28	26.90
utter-project/EuroLLM-9B-Instruct	45.21	49.79	48.68	50.00	49.60	52.27	54.44	52.86	53.15	50.58	51.91	47.14

ABOUT

LLMzSzŁ (LLM-y za Szkolną Ławą, LLMs Behind the School Desk) is a comprehensive LLM benchmark for Polish, based on a collection of Polish national exams, including both academic and professional tests extracted from the archives of the Polish Central Examination Board.

The benchmark was prepared by Adam Mickiewicz University's Center for Artificial Intelligence.

$$ \begin{array}{|c|c|c|c|} \hline \textbf{Exam} & \textbf{Discipline} & \textbf{Number of Questions} & \textbf{Original Number of Questions} \\ \hline 8th-Grade Exam & Polish Language & 15 & 101\\ 8th-Grade Exam & Mathematics & 35 & 99\\ Middle School Exam & Mathematics & 110 & 304\\ Middle School Exam & Nature & 65 & 226\\ High School Exam & Biology & 21 & 169\\ High School Exam & Physics & 136 & 399\\ High School Exam & Mathematics & 220 & 492\\ Professional Exam & Arts & 1778 & 3480\\ Professional Exam & Mechanical, Mining and Metallurgical & 9097 & 30240\\ Professional Exam & Agriculture and Forestry & 7344 & 23200\\ \hline \end{array} $$

Copy the following snippet to cite these results

@misc{eval-harness,
      title        = {A framework for few-shot language model evaluation},  
      author       = {Gao, Leo and Tow, Jonathan and Abbasi, Baber and Biderman, Stella and Black, Sid and DiPofi, Anthony and Foster, Charles and Golding, Laurence and Hsu, Jeffrey and Le Noac'h, Alain and Li, Haonan and McDonell, Kyle and Muennighoff, Niklas and Ociepa, Chris and Phang, Jason and Reynolds, Laria and Schoelkopf, Hailey and Skowron, Aviya and Sutawika, Lintang and Tang, Eric and Thite, Anish and Wang, Ben and Wang, Kevin and Zou, Andy},
      year         = 2023,
      publisher    = {Zenodo},
      doi          = {10.5281/zenodo.10256836},
      url          = {https://zenodo.org/records/10256836}
},
@inproceedings{gralinski-2019-geval,
    title = "{GE}val: Tool for Debugging {NLP} Datasets and Models",
    author = "Grali{'n}ski, Filip  and
      Wr{'o}blewska, Anna  and
      Stanis{\l}awek, Tomasz  and
      Grabowski, Kamil  and
      G{'o}recki, Tomasz",
    year = "2019",
    doi = "10.18653/v1/W19-4826",
    url = "https://aclanthology.org/W19-4826/",
}
}

If this work is useful to you, please consider citing:

@misc{jassem2025llmzszl,
      title={LLMzSz{\L}: a comprehensive LLM benchmark for Polish}, 
      author={Krzysztof Jassem and Michał Ciesiółka and Filip Graliński and Piotr Jabłoński and Jakub Pokrywka and Marek Kubis and Monika Jabłońska and Ryszard Staruch},
      year={2025},
      eprint={2501.02266},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2501.02266}, 
}

Built with Gradio logo

Adam Mickiewicz University'sCenter for Artificial Intelligence

LLMzSzŁ: a comprehensive LLM benchmark for Polish

Overall scores

Middle School Exam

Subjects Covered

Recent Changes

8th-Grade Exam

Subjects Covered

Purpose and Role

High School Exam

Mandatory Subjects

Additional Subjects

Minimum points required

Professional Exam

Purpose and Importance

ABOUT

Adam Mickiewicz University's
Center for Artificial Intelligence