DeepSeek open-sources DeepInfer framework to cut AI inference costs, targets Nvidia and Huawei GPUs

DeepSeek released DeepInfer, an open-source inference framework optimized for its V4.1-Flash model on Nvidia SM100 and Huawei GPUs, aiming to reduce AI deployment costs. A senior researcher framed the open vs. closed-source AI divide as a battle between 'light and darkness', signaling intensifying competitive rhetoric. The move could pressure closed-source rivals by lowering entry barriers, though actual cost savings and adoption remain unverified.

AI disclosure: Model-written summary of the source, not a verified fact. Check the sources. How Mosaic works