DeepSeek updates its R1 reasoning AI model, releases it on Hugging Face

China Startup DeepSEK published an updated version R1 Reasoning AI model on The developer platform embraces the face After announcing Wednesday morning in a Wechat message.

The updated R1 under the Permisive MIT license is a “small” update for Deepseek’s WECHat announcement that it can be used. Hugging Face Repository does not have a description of the model – internal components of a model that directly direct configuration files and weights, behaviors.

The size weighs 685 billion in parameters, the updated R1 is quite large. (“Settings” is synonymous with “Weights”.) Without modification, the model may not work on consumer level hardware.

DeepSEek rose to the popularity earlier this year An escape for money after the release of R1, which gives models from Openai. There is a start it was raised Some regulators are some regulators that say that Deepseek’s technology violates the risk of national security.