To evaluate the roles between two different HIV self-testing models in promoting HIV-testing among men who have sex with men (MSM). This paper focuses on two HIV self-testing service models. The first; is the online self-testing model (HIV self-testing conventional model) with the sexual health promotion network platform. The other one is an innovative HIV self-testing model (secondary distribution model), based on the previous program. The two different self-testing models, including the number of indexes and alters, the positive rate, and the demographics of indexes and alters, are compared. The influence of volunteers with or without leadership on the number of HIV self-test kits distributed or self-use is analyzed through the leadership survey scale. The return rates of HIV self-testing results in the two models are 94.7%(323/341) and 99.2%(1 141/1 150), respectively, within 30 days. The proportion of alters in the secondary distribution is significantly higher (45.9%,281/612) than the conventional HIV self-testing (6.3%,20/318). In the secondary distribution model, the difference between the number of indexes and alters indicators including age, marital status, residence, sex orientation, anal sex with men in the past six months, and HIV test are statistically significant ( test, all <0.05). The opinion leader of MSM has significantly impacted the promotion of HIV self-testing (<0.05). Both models can promote HIV self-testing, result return, and HIV positive detection among MSM. In terms of expanding the testing and detection of HIV positive, the secondary distribution mode shows more obvious advantages, which significantly promotes a large number of MSM who have never been tested for HIV to undergo HIV testing. Influential indexes have a significant effect on increasing the HIV testing rate and promoting HIV testing among MSM.