PANews reported on March 26 that according to DeepSeek's official announcement, the V3-0324 model has been released, focusing on improving reasoning capabilities, front-end development skills, and tool calling intelligence, and is suitable for conventional reasoning tasks. This version continues the MIT open source agreement, and the model weights have been simultaneously launched on the Hugging Face platform.
DeepSeek released V3-0324, significantly improving reasoning performance and tool usage capabilities
Share to:
Author: PA一线
This content is for informational purposes only and does not constitute investment advice.
Follow PANews official accounts, navigate bull and bear markets together
Recommended Reading
