China’s DeepSeek Unveils AI Model Halving Costs by 50% – The ‘Sparse Attention’ Revolution
Hangzhou-based DeepSeek burst onto the AI scene earlier in 2025 with its R1 model techcrunch.com. This time, DeepSeek’s announcement focuses on efficiency. On Sept. 29 the company published a post unveiling DeepSeek-V3.2-Exp, an experimental large language model built on its V3 series techcrunch.com techcrunch.com. According to DeepSeek, V3.2-Exp maintains similar reasoning performance to V3.1 but uses far less compute for long inputs. The key innovation is a “DeepSeek Sparse Attention” mechanism: rather than comparing every token to every other in a long document, DSA first uses a “lightning indexer” to pick out important excerpts, then a fine-grained selector to zoom in on the most salient words inside them techcrunch.com hindustantimes.com. This two-stage pruning means the model can “handle a large amount of data” more cheaply, processing tens of thousands of tokens without exploding costs techcrunch.com venturebeat.com. DeepSeek’s announcement on Hugging Face explicitly calls V3.2-Exp an “intermediate step toward our next-generation architecture” reuters.com. In practice, it built V3.2-Exp by adding DSA on top of its V3.1-Terminus model venturebeat.com. The company also released the full model weights and code under an open-source license on Hugging Face and GitHub techcrunch.com venturebeat.com, continuing its commitment to transparency. As VentureBeat notes, anyone can now download,
