News on October 28: Recently, observant netizens noticed that DeepSeek had once again changed to a brand-new large-model fingerprint, sparking widespread speculation that it is about to release a major new model. At present, entirely new code including DeepSeek-V5, V5-open, and V5-next has officially appeared in the official DeepSeek Harness code repository.
Judging from the code structure, the current V5-related content mainly serves as a placeholder, but this clearly indicates that the release mechanism for the new model is ready. When it is officially launched in the future, the official team will only need to modify the /models configuration on the server side to push V5 directly, without forcing a client-side release. Specifically, the code system covers two completely different delivery paths: deepseek-v5-open follows a no-version-threshold route, while deepseek-v5-next is a path that requires the client to be upgraded.
Of course, the existing code only proves that DeepSeek has laid the groundwork for the release of the V5 series of large models; it does not mean that the series will be launched immediately. Looking back at its recent pace, although V4.1Flash, released on September 17, adopted a completely new architecture, its naming still followed the 4.0 series. Considering the feedback when the previous generation's official Pro version was released, it is also a reasonable inference that the official team may directly name the next-generation flagship large model DeepSeek V5Pro in the future.
Industry analysts believe that, given that the Flash version was released not long ago, the V5Pro series is unlikely to go online immediately within October and is expected to make its official public debut at the earliest in November. If progress goes smoothly, V5Pro and V5Flash may be updated together in mid-November, making the product naming lineup clearer and more straightforward.