According to CoinMeta and OneMillion.AI, a Spanish company Multiverse Computing has promoted its model Quasar with 438B parameters as “the strongest AI model in Europe.” However, the community discovered a close relationship between this model and Zhipu GLM-5.2. The community has labeled this model as “Quasar 438B (including max, based, on, GLM-5.2),” and pointed out that it is based on GLM-5.2. On August 5th, when Quasar 438B was launched on beta, the update logs showed that its capabilities were “completely identical” to those of GLM-5.2. The community questions whether Quasar is merely a compressed or modified version of GLM-5.2 with a new name. This model has a total of about 753B parameters; when 40B of these parameters are activated for reasoning, its intelligence index scores 53 points. After reducing the parameters to 438B, the score drops to 43 points, but the output speed is approximately 174 token per second, which is significantly faster than GLM-5.2’s speed of about 68 token per second.