Objective To investigate changes in serum levels of key pyroptosis-related factors， including gasdermin D（ GSDMD）， nucleotide-binding oligomerization domain-like receptor protein 3（ NLRP3）， interleukin（IL）- 1β， and IL-18 in patients with acute large artery atherosclerotic（ LAA） cerebral infarction， and to establish a predictive model for poor prognosis combining with clinical indicators. Methods A total of 223 patients with acute LAA cerebral infarction admitted to the First Affiliated Hospital of Henan Medical University between June 2023 and October 2024 were included in the study. According to the scores on the 3-month Modified Rankin Scale（ mRS）， patients were divided into a good prognosis group（ mRS≤ 2； n=150） and a poor prognosis group（ mRS ＞ 2； n=73）. Enzyme-linked immunosorbent assay（ ELISA） was used to detect the serum levels of GSDMD， NLRP3， IL-1β， and IL-18 in patients at admission and on day 7 of admission. Spearman rank correlation was used to analyze the correlation between various biomarkers and the National Institute of Health Stroke Scale（ NIHSS） score at admission and 3-month mRS score. Binary Logistic regression was used to analyze independent risk factors for poor prognosis. After adjusting for potential confounding factors such as age and coronary heart disease， a nomogram prediction model was constructed， and its predictive performance was evaluated using the receiver operating characteristic（ ROC） curve and the Hosmer-Lemeshow test. Results In the poor prognosis group， the GSDMD， NLRP3， IL-18 levels at admission 1 249.82（ 1 071.14， 1 444.40） pg/ml， 1 297.44（ 1 075.45，1 392.13） pg/ml， and（ 173.04±28.32） ng/ml， respectively， on day 7 of admission， the GSDMD， NLRP3， and IL-18 levels were（ 1 275.01±140.79） pg/ml， （1 264.45±216.49） pg/ml， and 49.65（ 48.84， 50.64） pg/ml， respectively， all of which were higher than those in the good prognosis group［ 1 109.25（ 956.85， 1 265.67） pg/ml， 1 160.55（ 911.05，1 346.26） pg/ml， （163.66±20.87） ng/ml， and（ 1 223.37±174.90） pg/ml，（ 1 137.43±237.93） pg/ml， and 43.68（ 38.84， 47.32） pg/ml］， all of which were statistically significant（ Z/t=-2.413，-2.597， -2.517，-2.368， 3.851， -7.719； all P＜0.05）. The difference in the change in serum IL-1β level between the two groups at admission and on day 7 of admission was statistically significant（ Z=-4.034， P ＜ 0.001）. Patients' serum levels of GSDMD， NLRP3， IL-1β， and IL-18 at admission， as well as GSDMD and IL-1β levels on day 7 of admission， were all positively correlated with the NIHSS score at admission（ r=0.344， 0.227， 0.153， 0.180， 0.149， 0.169； all P＜ 0.05）； serum GSDMD and NLRP3 levels at admission， as well as IL-1β levels on day 7 of admission， were all positively correlated with the 3-month mRS score（ r=0.201， 0.150， 0.367； all P＜ 0.05）， and all differences were statistically significant. Binary Logistic regression analysis showed that a high percentage of neutrophils at admission［ OR=1.073， 95%CI（ 1.007，1.143）， P=0.029）， a high NIHSS score at admission［ OR=1.584， 95%CI（ 1.316，1.907）， P＜0.001）， and a high GSDMD level at admission［ OR=1.004， 95%CI（ 1.001，1.006）， P=0.013］， and high IL-1β level on day 7 of admission［ OR=1.438， 95%CI（ 1.254，1.649）， P＜0.001］ were statistically independent risk factors for poor 3-month prognosis in patients with LAA cerebral infarction. The area under the ROC curve（ AUC） of the combined nomogram model for predicting poor prognosis was 0.959 ［95%CI（ 0.932，0.986）］， with a sensitivity of 89.04% and a specificity of 92.00%. The Hosmer-Lemeshow goodness-of-fit test showed good calibration of the model（ χ2=4.500，P=0.809）. Conclusions Elevated serum levels of GSDMD and IL-1β in patients with LAA cerebral infarction are positively correlated with poor prognostic outcomes， and a combined predictive model constructed by combining the baseline NIHSS score can improve the accuracy of prognostic assessment.