DeepSeek is in early talks for a new funding round at a $71 billion pre-money valuation.

Released DSpark speculative decoding framework

Began developing custom inference ASICs to cut GPU dependency

China's domestic AI chip budget share rises to 46%, driven by DeepSeek and Zhipu exploring custom silicon

Meituan open-sourced LongCat-2.0, a 1.6 trillion-parameter LLM trained entirely on domestic chips

Raised $7B in first major funding round, abandoning no-funding pledge

Plans to launch Harness AI coding agent team to compete with Claude Code

Published research paper on MBGR, a generative recommendation framework for multi-business platforms