DeepSeek updates R1 reasoning AI model
Chinese AI startup DeepSeek has quietly launched an updated version of its flagship R1 reasoning model, making it available to developers through the popular open-source platform Hugging Face. The update was announced in a WeChat post Wednesday morning.
Minor Upgrade, Major Model
According to DeepSeek, the new release is a “minor” upgrade over the original R1, but it’s still a heavyweight in the AI world. The updated R1 model weighs in at a massive 685 billion parameters, placing it among the largest publicly released AI models. Due to its size, it’s unlikely to be runnable on standard consumer hardware without significant modifications or distributed compute support.
Open for Business
The model is released under the permissive MIT license, meaning developers and companies alike can freely use, modify, and integrate R1 into commercial applications. While the Hugging Face repository currently lacks detailed documentation, it includes the essential configuration files and model weights required for deployment.
Strategic and Political Implications
DeepSeek gained widespread attention earlier this year with the launch of the original R1 model, which was touted as a potential rival to offerings from OpenAI. The company’s rapid rise has not gone unnoticed—particularly by U.S. regulators who have expressed concerns about the potential national security implications of such advanced technology being developed and shared openly from China.
Despite the political scrutiny, DeepSeek continues to push forward with open-access AI research, offering powerful tools to a global developer community.
With this release, DeepSeek further cements itself as a major player in the global AI race, leveraging transparency and accessibility to compete with some of the most advanced tech firms in the world.
