Background This study aimed to elucidate the treatment outcomes and prognosis of angioimmunoblastic T-cell lymphoma (AITL) patients in a real-world setting.Objectives The clinical efficacy of new drug applications was evaluated, alongside the predictive accuracy of prognostic models, to inform future research.Methods In this study, 82.9% of patients received a CHOP-like regimen, while 36.4% also received chidamide. We assessed the prognostic models' predictive power using the Cox proportional hazards model and concordance index (C-index).Results The median age of the patients in this study was 62.0 years, with 2-year progression-free survival (PFS) and overall survival (OS) rates of 36.1% and 60.3%, respectively. Complete response (CR) rates in first-line treatments were 21.6% for the chidamide-containing group and 28.1% for the chidamide-free group. The AITL scores, PIAI scores, and Chinese AITL scores demonstrated superior C-index values, with the Chinese AITL score providing the most distinct risk stratification. Advanced age (over 70 years), bone marrow involvement, and CD7 negativity were identified as significant prognostic factors associated with poorer PFS in both univariate and multivariate analyses. A novel prognostic model, the South China AITL Score, was constructed by combining these three factors, stratifying patients into low-risk and high-risk groups, with 5-year PFS rates of 81.5% and 34.6%, respectively. This model was successfully validated in an independent cohort.Conclusions The prognosis of AITL in real-world settings is poor, and the addition of chidamide did not show improvement in remission rates or survival. Our novel prognostic model, along with the Chinese AITL score, may enhance the identification of Chinese patients at varying risks for chemotherapy. Furthermore, the pathological marker CD7 is anticipated to emerge as a significant biomarker for the prognostic evaluation of AITL.