Added new model: deepseek-v3-0324deepseek-v3-0324 is the latest high-performance large language model launched by Novita, representing the newest version of the flagship conversational model series from the DeepSeek team. It adopts a Mixture-of-Experts (MoE) architecture and boasts a parameter scale of 685 billion parameters.Updated time: March 26, 2025
Optimized third-party login integrationFixed an issue where multiple clicks on third-party login prevented users from signing in.Updated time: March 28, 2025
Exclusive discounts now displayed in “Explore-Template andGPU Server” sectionCustomers with exclusive discounts now see discounted prices directly within the “Explore-Template and GPU Server” sections, rather than solely in billing. This provides clearer discount visibility to support decision-making.Updated time: March 26, 2025
Improved user interaction for the Model LibraryClicking “See All” in a Model Library category now takes users directly to a dedicated page showcasing all templates within that category, instead of redirecting them to the homepage.Updated time: March 24, 2025