Moonshot AI Closes ~$2B Funding Round at $20B Valuation
Beijing-based Moonshot AI, creator of the open-weight Kimi large language model series, has finalized a new financing round worth approximately $2 billion, valuing the startup at around $20 billion. The transaction was spearheaded by Long-Z Investments—the venture capital subsidiary of Meituan—with additional backing from China Mobile, CPE Yuanfeng, and other institutional investors.
Established in 2023 by Yang Zhilin, a researcher with prior experience at Meta AI and Google Brain, Moonshot has accumulated roughly $1.9 billion in capital commitments over the last six months. This latest valuation represents more than a two-fold increase from early 2025, underscoring intensifying global competition for accessible, open-source AI infrastructure.
Prithvi Geospatial AI Achieves First In-Orbit Deployment
A collaborative team from Adelaide University and SmartSat CRC has successfully operated NASA and IBM’s open-source Prithvi Geospatial foundation model on two distinct space-based platforms: the South Australian government’s Kanyini satellite, and the IMAGIN-e experimental payload hosted aboard the International Space Station.
The model, trained using over a decade of multispectral Earth observation imagery, enables rapid identification of flooding events, cloud cover assessment, and related geospatial analytics. Its compressed architecture maintained inference accuracy across both orbital environments, establishing a benchmark for deploying open-source AI capabilities in space missions.
Quantum Motion Raises $160M to Scale Silicon Quantum Processors
London-headquartered quantum computing firm Quantum Motion has secured $160 million in Series C funding, with DCVC and Kembara serving as lead investors. The company’s core innovation lies in silicon spin qubits fabricated via conventional CMOS processes—the same manufacturing foundation underpinning today’s classical microprocessors.
This design philosophy unlocks a critical scalability pathway: high-volume production can utilize existing semiconductor foundries, eliminating the need for dedicated facilities or bespoke fabrication tooling. The approach positions Quantum Motion to accelerate development of error-corrected, large-scale quantum processors.