DeepSeek's new model slashes AI inference costs
DeepSeek unveils V3.2-exp, a model designed to significantly reduce inference costs with its innovative sparse attention system.
posted on 09/29/2025DeepSeek unveils V3.2-exp, a model designed to significantly reduce inference costs with its innovative sparse attention system.
posted on 09/29/2025The Trump administration is considering restricting Chinese AI lab DeepSeek's access to Nvidia chips and US market amid IP concerns.
posted on 04/17/2025