DeepSeek has unveiled its V3.2-Exp AI model, featuring breakthrough sparse attention for efficient processing of long texts. The company also slashed API prices by over 50%, intensifying global AI competition. V3.2-Exp matches predecessor performance with higher efficiency and is available via app, web, and open-source platforms. #AI #DeepSeek #SparseAttention #TechInnovation #ModelRelease
DeepSeek's new AI model paves way for its next-gen architecture
Chinese AI developer, DeepSeek, has launched an experimental release of a new model, touted as an efficient and advanced step towards next-generation architecture.
https://www.newsbytesapp.com/news/science/china-s-deepseek-launches-new-ai-model-slashes-api-prices/story
2 days ago
