Baidu, Tencent, and Zhiyu AI Invest in Wulian Chip


Tencent recently open-sourced the compact AI translation model Hy-MT1.5-1.8B-1.25bit, supporting 33 languages, 5 dialects, and 1056 translation directions, fully offline on smartphones. Using 1.25-bit quantization, it aggressively compresses from 3.3GB while maintaining high performance, having won 30 international machine translation competition championships.....
Tencent's intelligent workbench ima launched the personal knowledge agent 'copilot' on April 29, upgrading AI search into an evolving knowledge partner through deep personalized memory and full-scenario perception. Its core is an autonomous evolving memory system integrating four modules: persona, user profile, long-term memory, and experience skills, marking AI's transition from a single-dialogue tool to a long-term intelligent agent.....
AI unicorn DeepSeek has launched its first external fundraising round, targeting a valuation over $20 billion, with Tencent and Alibaba expressing interest. Despite relying on founder Liang Wenfeng's quantitative trading funds and having no urgent need for capital, the move aims to retain key talent through equity incentives amid recent departures of core members like R1 paper author Guo Daya and researcher Wang Bingxuan.....
Tencent released and open-sourced the new AI model 'Hunyuan Hy3 Preview', the most intelligent in its series. Upgrades cover complex reasoning, instruction following, contextual learning, code processing, and agents. It uses a hybrid expert architecture combining fast and slow thinking, with 295 billion parameters, to enhance overall performance and intelligence.....
Tencent launches open-source AI model Hy3Preview with 29.5B parameters, supporting long contexts and improved performance in reasoning, instruction following, and code processing, emphasizing practicality, real-world evaluation, and cost-effectiveness.....