Nvidia is bolstering hardware optimization for Chinese open-source AI models including DeepSeek and Alibaba's Qwen, seeking to maintain its position in the global AI ecosystem despite growing concerns from US lawmakers about supporting Chinese technology development.
Anthropic reports that Claude Code successfully handled daily software maintenance tasks with a 46% merge rate for its pull requests, while DeepSeek launched DeepSeek Harness as an open source alternative alongside its V4-Pro model update.
California-based Applied Compute, which helps companies develop and deploy open-weight AI models, is in talks to raise hundreds of millions of dollars at approximately a $3 billion valuation in a funding round led by investor Elad Gil.
Hugging Face CEO Clément Delangue told CNBC that Chinese developers now account for 41% of AI model downloads on the platform, the largest share globally, as China gains ground in open-source AI development. The announcement comes amid a broader debate about U.S.-China AI competition and prompted Meta CEO Mark Zuckerberg to call for reduced U.S. regulatory barriers on open-source AI.
Meta Superintelligence Labs released Muse Spark, its first frontier AI model and first without open weights, marking a significant shift from Meta's open-source strategy. The model now powers Meta AI app and Meta.ai website.
Google's DeepMind released Gemma 4, a family of four open-weight AI models ranging from edge devices to workstations, marking the first time Google has used the fully open Apache 2.0 license instead of its previous restricted licensing. The models support up to 256K context windows and 140+ languages.