DeepSeek seeks ‘total victory’ over closed-source AI with new open-source tool
Chinese artificial intelligence start-up DeepSeek has fully open-sourced an internally developed framework designed to run AI models efficiently, as a senior researcher at the company described the growing divide between open-source and closed-source AI as a battle between “light and darkness”. The company on Tuesday released DeepInfer, an inference framework tailored to support its…
Chinese AI startup DeepSeek has released DeepInfer, an inference framework specifically designed to optimize its DeepSeek-V4.1-Flash model for both Nvidia's SM100 Blackwell GPUs and Huawei's Ascend 950 chips. The company claims the release of DeepInfer will lower compute costs, reduce energy usage, and make AI more affordable for users.
DeepSeek-V4.1-Flash, launched in September, boasts a new architecture that reduces memory requirements while outperforming its predecessors. The open-source framework allows third-party cloud providers to serve the model on their infrastructure, leading to potential revenue gains for the models' users. DeepSeek senior researcher Chen Deli described the open-source vs. closed-source AI debate as a battle between "light and darkness," emphasizing that his company's goal is "total victory."
This release comes amid growing competition between Chinese AI companies and leading American rivals like Anthropic and OpenAI, who promote closed-source AI models. Despite criticism from US industry figures, Chinese firms continue to release open-weight AI models, with Beijing maintaining support for open-source AI adoption globally.
Written by urgent.news from SCMP Tech's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
